framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,10240,7.548380533854167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,12288,8.948833889431423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,16384,11.654150221082899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,10240,12.824673122829862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,12288,15.342364841037325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,8192,5.967103322347005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,8192,10.29634518093533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,16384,20.217741224500866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,7168,5.292071024576823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,7168,8.981141832139757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,6144,4.574204338921441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,6144,7.644284566243489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,5120,6.422179751925999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,5120,5.0516916910807295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,4096,5.158511267768012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,4096,3.717016855875651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,12288,10.6793941921658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,10240,8.922033521864149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,3584,4.545988294813368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,3584,3.245768017239041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,3072,3.9026035732693143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,3072,2.7779057820638022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,5120,4.718756357828776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,16384,15.909542507595487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,2560,3.279150221082899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,2560,2.3113636440700955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,3584,3.3171876271565757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,2048,1.8540791405571833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,6144,5.525569068060981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,2048,2.679596371120877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,1536,1.3855689366658528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,1536,2.035747528076172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,1024,1.398232036166721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,1024,0.9497564103868273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,3072,2.8364043765597873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,768,1.1376337475246854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,768,0.7152906523810493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,8192,7.396338568793403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,2560,2.310426712036133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,512,0.5433831214904785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,512,0.8229680061340332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,256,0.5256657600402832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,4096,3.798592037624783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,256,0.48427735434638125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,128,0.4424168798658583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,128,0.437624004152086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,1536,1.5073724322848852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,2048,1.9657164679633246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,64,0.41037776735093856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,64,0.40862133767869735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,32,0.3988613287607829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,32,0.40238043997022843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,256,0.5160693592495388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,512,0.6347289085388184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,1024,1.0478319591946073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,768,0.8397262361314562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,7168,6.404923756917317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,16384,3.971082475450304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,12288,4.015231238471137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,16384,5.233559926350911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,12288,3.040913687811958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,65536,18.076778835720486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,10240,2.561977810329861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,10240,3.301354726155599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,8192,2.695156521267361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,8192,2.0197830200195312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,7168,2.33552614847819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,16384,4.866182115342882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,65536,21.228198581271702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,7168,1.7955955929226344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,6144,1.564152929517958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,6144,2.0615066952175565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,5120,1.7196213404337566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,12288,3.713314480251736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,5120,1.5098275078667536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,4096,1.3446444405449762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,10240,3.0871404012044272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,4096,1.1963599522908528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,3584,1.2107342614067924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,3584,1.0486169391208224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,8192,2.4649270375569663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,3072,1.0229270723130968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,3072,0.9158684412638346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,65536,20.126937866210938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,128,0.4740533298916287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,2560,0.8170017666286893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,7168,2.1369120279947915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,2560,0.7610159979926215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,2048,0.648579544491238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,2048,0.6109333568149143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,6144,1.8341183132595484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,1536,0.46105157004462344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,1536,0.4977368778652615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,1024,0.34717689620123965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,1024,0.3112124337090386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,5120,1.5520657433403864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,768,0.3007448779212104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,768,0.23361600769890678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,4096,1.2402755949232314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,512,0.20887733830346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,3584,1.1070364846123588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,512,0.1708142227596707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,256,0.12661777602301702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,256,0.1381208896636963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,3072,0.9334942499796549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,128,0.11504444811079238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,128,0.12133688396877712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,2560,0.8006141980489095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,64,0.10713599787818061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,64,0.11901777320437962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,32,0.10395289129681057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,32,0.11815110842386882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,1536,0.5124399926927354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,2048,0.6543813281589085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,768,0.3009173340267605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,256,0.1650408903757731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,1024,0.37229688962300617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,512,0.23274755477905273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,16384,3.804071214463976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,16384,3.387619654337565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,12288,2.917687098185221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,12288,2.5207174089219833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,10240,2.4500569237603083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,10240,2.134660508897569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,8192,1.9742452833387587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,65536,15.060111151801216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,8192,1.7023688422309027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,7168,1.7542604870266383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,7168,1.512183083428277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,65536,15.063693576388888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,16384,4.143302917480469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,6144,1.4829609129163954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,6144,1.324841817220052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,12288,3.103748533460829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,5120,1.24181334177653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,5120,1.2280471589830186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,4096,0.9928959740532769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,10240,2.669135199652778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,4096,0.9888898001776801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,3584,0.835901313357883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,3584,0.8725546730889214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,65536,16.909700181749134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,8192,2.0570062001546225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,3072,0.75436888800727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,3072,0.7051368819342719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,2560,0.5710684458414713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,2560,0.6305564244588217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,7168,1.8028142717149522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,2048,0.4773724344041612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,2048,0.5060177909003364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,6144,1.5211066140068903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,1536,0.36687999301486546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,1536,0.3810231155819363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,128,0.14800177680121526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,5120,1.2878292931450737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,1024,0.2564648787180583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,1024,0.26185067494710285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,768,0.20767822530534533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,768,0.19535822338528105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,4096,1.0609369277954102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,3584,0.9304595523410373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,512,0.15336444642808703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,512,0.14613866806030273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,256,0.09806755516264175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,256,0.11317511399586995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,2560,0.6737164391411675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,128,0.08679288625717163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,128,0.09887555572721694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,3072,0.8007893562316895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,64,0.08003733555475871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,64,0.10527288913726807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,32,0.07869688669840495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,32,0.09408444166183472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,2048,0.5550319883558485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,1536,0.4334079954359267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,256,0.13797067271338567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,512,0.19374933507707384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,768,0.2527066601647271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,16384,3.01092169019911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,1024,0.3143920103708903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,16384,3.410795635647244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,12288,2.3190960354275174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,12288,2.6163092719184027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,10240,1.9244187672932942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,65536,12.914538065592447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,10240,2.142786661783854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,8192,1.5410346984863281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,8192,1.7244880464341905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,7168,1.5251057942708333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,65536,13.375047471788195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,7168,1.364850680033366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,6144,1.230979601542155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,6144,1.1874888737996419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,16384,3.7471110026041665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,5120,1.0881901846991646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,5120,1.1128418180677626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,12288,2.811929702758789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,4096,0.8530871073404948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,4096,0.8874764442443848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,10240,2.348298602634006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,8192,1.8482078976101344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,3584,0.7787671089172363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,3584,0.7151306470235189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,65536,15.175799899631075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,7168,1.6132657792833116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,3072,0.5926408767700195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,3072,0.6745591163635254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,2560,0.5637306637234157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,2560,0.5171519915262858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,2048,0.4045111073387994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,2048,0.4579937722947862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,128,0.13301422860887316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,6144,1.4026319715711806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,1536,0.31236444579230416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,1536,0.34346310297648114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,1024,0.2227768898010254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,1024,0.23373066054450142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,5120,1.1794186698065865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,768,0.1791884501775106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,768,0.1757048898273044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,4096,0.9486302269829644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,512,0.1365928914811876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,512,0.12609067228105333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,3584,0.8497226503160266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,256,0.094014220767551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,256,0.09799466530481975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,2560,0.6146088706122504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,128,0.07487999730639987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,128,0.08561955557929145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,3072,0.7214613490634494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,64,0.069069332546658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,64,0.0823413332303365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,32,0.06835022237565783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,32,0.08193511433071561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,2048,0.5042719841003418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,1536,0.3942248821258545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,1024,0.28685156504313153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,768,0.22906221283806694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,512,0.1740328868230184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,256,0.12311200300852458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,16384,2.680614259507921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,16384,2.735776053534614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,12288,2.1425467597113714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,12288,2.0566399892171225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,65536,11.274898105197481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,10240,1.7146915859646266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,10240,1.7561670939127605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,8192,1.3716906441582575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,8192,1.4193181991577148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,7168,1.1341209411621094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,65536,10.875406901041666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,7168,1.2179608874850802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,6144,1.059559080335829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,6144,1.0596862369113498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,16384,3.3358436160617404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,12288,2.5003723568386502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,5120,0.8562791082594129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,5120,0.9702178107367622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,65536,13.367296006944445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,4096,0.6213022338019477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,4096,0.7716897858513726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,10240,2.0766595204671225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,3584,0.5521644486321343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,3584,0.6749706798129611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,8192,1.7030533684624567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,3072,0.48389424218071836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,3072,0.5692470868428549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,2560,0.4028328789605035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,2560,0.4745608965555827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,7168,1.4558284547593858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,2048,0.3332364559173584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,2048,0.37500890096028644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,6144,1.2541920344034831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,1536,0.2612088786231147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,1536,0.2809635533226861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,128,0.10767466492123073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,1024,0.18302044603559706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,1024,0.19120444191826713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,5120,1.0563217798868816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,768,0.14131822850969103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,768,0.1429822180006239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,3584,0.7521600193447537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,512,0.1060604453086853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,512,0.10844533310996161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,4096,0.8539546860588921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,256,0.06756444772084554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,256,0.07991821898354425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,2560,0.5613457891676161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,128,0.05947644180721707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,128,0.06899199883143108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,3072,0.6508391168382432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,64,0.05499733487764994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,64,0.06530577606625028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,32,0.05411200059784783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,32,0.06681778033574422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,1536,0.3573511176639133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,2048,0.4567519823710124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,256,0.10830310980478923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,512,0.15345422426859537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,1024,0.2561075687408447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,768,0.20389333036210802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,16384,2.514322704739041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,16384,2.3061421712239585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,65536,10.659203423394098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,12288,1.7320409350925023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,12288,1.9220595889621312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,65536,9.00455050998264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,10240,1.4052471584743922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,10240,1.6083022223578558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,8192,1.2928319507175021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,8192,1.1305635240342882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,7168,0.9255350960625542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,7168,1.145352045694987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,6144,0.8447244432237414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,6144,0.9932809405856662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,16384,3.135649787055122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,12288,2.3337813483344183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,5120,0.6836098035176595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,5120,0.899902237786187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,10240,1.9618036482069228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,65536,12.780815972222221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,4096,0.7182337972852918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,4096,0.5323706732855903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,8192,1.563351101345486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,3584,0.6270400153266059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,3584,0.46581421958075625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,3072,0.5360142389933268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,3072,0.4032328923543294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,2560,0.34034132957458496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,2560,0.4448249075147841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,7168,1.3779946433173285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,2048,0.27758754624260795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,2048,0.35451732741461855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,6144,1.1857866711086698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,1536,0.21559377511342367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,1536,0.2673715485466851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,128,0.09267199701733059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,1024,0.15363910463121203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,1024,0.18033688598208955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,3584,0.7121484544542102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,768,0.12431200345357259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,768,0.13787378205193415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,5120,1.0003111097547743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,512,0.09395111269421047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,512,0.10243555572297837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,2560,0.524702231089274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,4096,0.8077911270989312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,256,0.06015021933449639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,256,0.07610666751861572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,128,0.05373422304789225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,128,0.06370221906238131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,3072,0.621089776357015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,64,0.049775110350714795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,64,0.06100710895326403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,32,0.04801511102252536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,32,0.06121066543791029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,1536,0.3400142192840576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,2048,0.43372000588311094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,256,0.1005057758755154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,1024,0.24383465449015299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,512,0.1454533338546753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,768,0.1921057833565606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,65536,9.758154127332899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,65536,7.785652160644531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,16384,2.366426679823134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,16384,1.999683592054579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,12288,1.804300520155165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,12288,1.5206435521443684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,10240,1.2510426839192708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,10240,1.5085084703233507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,8192,0.9914568795098199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,8192,1.206865734524197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,7168,1.0692542394002278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,7168,0.8694693247477213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,6144,0.7405573527018229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,6144,0.929750230577257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,65536,11.901880900065104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,16384,2.945272021823459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,5120,0.5935404565599229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,12288,2.2140657636854386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,5120,0.8389866616990832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,4096,0.4758897887335883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,4096,0.6689768897162544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,3584,0.4128382205963135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,3584,0.582488907708062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,10240,1.8300880855984156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,3072,0.36423201031155056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,8192,1.4678577846950953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,3072,0.5021146668328179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,2560,0.30272799068027073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,2560,0.41646220948961044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,7168,1.3148142496744792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,2048,0.2446826563941108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,2048,0.3291155497233073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,6144,1.1177084181043837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,1536,0.19044355551401773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,1536,0.24599377314249674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,128,0.09397688839170669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,1024,0.13806845082177058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,1024,0.16937243938446045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,3584,0.6761306656731499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,768,0.108551111486223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,5120,0.9435075124104818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,768,0.12752355469597712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,512,0.08161422279145983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,512,0.09487377934985691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,4096,0.7618044217427572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,256,0.05318489008479648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,256,0.06910666492250231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,2560,0.4975760247972276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,128,0.04776711265246073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,128,0.05825510952207777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,3072,0.5838106473286947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,64,0.04372888803482056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,64,0.054902222421434194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,32,0.04232711262173123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,32,0.05527377790874905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,1536,0.3181013266245524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,2048,0.4116284317440457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,256,0.09579466448889838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,65536,9.056466844346788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,512,0.1350195540322198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,128,0.0811413327852885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,65536,6.684628380669488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,16384,2.20180172390408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,16384,1.6391955481635199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,768,0.17899110582139757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,12288,1.2846861945258248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,12288,1.6770515441894531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,10240,1.0511307186550565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,10240,1.4019991556803386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,8192,1.1279751459757488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,8192,0.793774233924018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,1024,0.22812178399827746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,7168,0.9967404471503364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,7168,0.7244248920016818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,6144,0.8658924632602267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,6144,0.6086506843566895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,5120,0.47748978932698566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,5120,0.7701662381490072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,65536,11.293568929036459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,4096,0.3960533406999376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,4096,0.6214319864908854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,16384,2.7536328633626304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,12288,2.048458735148112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,3584,0.5391448868645562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,3584,0.34715999497307676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,10240,1.7259546915690105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,3072,0.31099287668863934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,3072,0.46593867407904727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,2560,0.25253955523173016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,8192,1.3937146928575304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,2560,0.38815644052293563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,2048,0.20991199546390107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,2048,0.3090088897281223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,7168,1.2254124747382271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,1536,0.16078044308556452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,1536,0.2311697800954183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,6144,1.0560933219061959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,1024,0.11477422714233398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,1024,0.15743554963005915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,5120,0.8804648717244467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,768,0.09305510918299358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,768,0.11873866452111138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,3584,0.634310245513916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,512,0.07123288843366835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,512,0.08821244372261895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,2560,0.4681697951422797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,256,0.048862222168180675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,256,0.06399910979800753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,4096,0.7242942386203342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,128,0.040129777457978986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,1536,0.3000248803032769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,128,0.05212177832921346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,64,0.03578755590650771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,64,0.04857422245873345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,32,0.03605866763326857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,32,0.049581332339180835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,3072,0.5504399935404459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,2048,0.38422666655646426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,768,0.16815111372205946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,16384,1.3421697616577148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,1024,0.21299289332495797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,16384,2.0482205284966364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,12288,0.8781706492106119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,65536,5.197986602783203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,65536,8.206893920898438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,12288,1.5563217798868816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,10240,0.7463537851969401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,512,0.12540711296929255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,128,0.07448710997899373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,10240,1.3054435518052843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,256,0.08676799800660874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,8192,0.6383040216233995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,8192,1.0375181833902996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,7168,0.5381235546535915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,7168,0.9183804194132487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,6144,0.7874719831678602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,6144,0.4484755727979872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,5120,0.376801782184177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,5120,0.7056408988104926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,16384,2.560145696004232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,4096,0.310240003797743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,4096,0.5673679775661892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,12288,1.9500817192925348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,3584,0.2735004425048828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,3584,0.4944000244140625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,65536,10.272413465711805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,3072,0.23948799239264595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,10240,1.6117235819498699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,8192,1.2874061796400282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,3072,0.4255857732560899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,2560,0.19816177421145967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,7168,1.143721792432997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,2560,0.3596204386817084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,2048,0.16476178169250488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,2048,0.2871955500708686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,1536,0.12665332688225642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,1536,0.21590399742126465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,6144,0.9804461797078451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,1024,0.09023555782106187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,1024,0.14450489150153265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,5120,0.8291315502590604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,768,0.07409955395592584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,768,0.10869599713219537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,4096,0.6708809004889594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,3584,0.5971840222676595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,512,0.05641333262125651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,512,0.07995644542906019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,256,0.03755022088686625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,256,0.055753780735863574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,2560,0.4356355667114258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,128,0.032849778731664024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,128,0.04573511083920797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,1536,0.2799004448784722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,64,0.029743111795849268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,64,0.043987555636299976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,32,0.02976622184117635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,32,0.042985777060190834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,3072,0.5153893364800347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,2048,0.3587191104888916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,768,0.15525510576036242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,1024,0.19883022043440077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,16384,1.1788880030314128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,16384,1.9562142690022786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,12288,0.7953591346740723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,65536,4.509654998779297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,65536,7.832791646321614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,12288,1.4880399703979492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,10240,0.6448613272772895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,256,0.07915111382802327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,10240,1.245913823445638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,128,0.06646311283111572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,8192,0.9931955337524414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,8192,0.5489946471320258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,512,0.1143360005484687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,7168,0.45685778723822695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,7168,0.8731395403544108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,6144,0.7536248630947537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,6144,0.40051288074917263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,5120,0.3315209017859565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,5120,0.6715182198418511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,16384,2.46344969007704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,4096,0.27261421415540904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,4096,0.543434672885471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,12288,1.8540977901882594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,3584,0.2416693369547526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,3584,0.47948355144924587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,65536,9.926363627115885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,10240,1.5562942292955186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,3072,0.21277688609229195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,8192,1.2454036076863606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,2560,0.17471555868784586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,3072,0.40818222363789874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,7168,1.0966338051689994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,2048,0.14370755354563394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,2560,0.3447102175818549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,2048,0.27648621135287815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,1536,0.11360355218251546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,1536,0.20755378405253092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,6144,0.9589182535807291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,1024,0.08115110794703166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,1024,0.14100978109571669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,5120,0.7994000116984049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,768,0.06576710939407349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,768,0.10535999801423813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,3584,0.5720204247368706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,512,0.05031555559900072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,512,0.07638044489754571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,4096,0.649794684516059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,256,0.03328355484538608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,256,0.05351288782225715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,2560,0.4257226785024007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,128,0.02951999836497837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,128,0.04327911138534546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,1536,0.26997600661383736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,64,0.027456889549891155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,64,0.04033333394262526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,32,0.026926222774717543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,32,0.03974577784538269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,768,0.14935110674964056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,3072,0.4979955355326335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,2048,0.3474408785502116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,16384,0.9024399651421441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,1024,0.1918355491426256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,16384,1.8809324900309246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,12288,0.6809013154771594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,65536,3.8640157911512585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,12288,1.4286773469712999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,65536,7.538033379448785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,10240,0.575310230255127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,10240,1.1978026496039498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,8192,0.4511466556125217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,256,0.07632711198594835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,128,0.061920887894100614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,8192,0.9533501731024848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,512,0.11044177744123672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,7168,0.841301335228814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,7168,0.4000302155812581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,6144,0.7261946466233996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,6144,0.3430639902750651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,5120,0.2890666590796577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,5120,0.6439475483364528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,16384,2.3829583062065973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,4096,0.23620088895161948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,4096,0.5194142129686143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,12288,1.7917315165201824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,3584,0.20991644594404432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,3584,0.4563724199930827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,65536,9.50459459092882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,3072,0.18843644195132783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,10240,1.5035173628065321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,3072,0.3923982249365912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,2560,0.15536710951063368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,2560,0.33041244082980686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,8192,1.2074453565809462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,2048,0.12673866748809814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,2048,0.2653191089630127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,7168,1.066298696729872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,1536,0.10036800305048625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,1536,0.20107111665937635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,6144,0.9173502392239041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,1024,0.07181688812043932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,1024,0.13623289267222086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,5120,0.773091581132677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,768,0.059009777175055615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,768,0.10079377889633179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,3584,0.5528915723164877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,512,0.04461866617202759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,512,0.07183555761973064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,4096,0.6297449005974664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,256,0.030488000975714788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,256,0.05006400081846449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,2560,0.4082079993353949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,128,0.026161778304311965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,128,0.04025955663786994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,1536,0.2605786588456896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,64,0.023894222246276006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,64,0.03757333424356248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,32,0.024988444315062627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,32,0.03687377770741781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,3072,0.4802568753560384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,768,0.14346667130788168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,2048,0.33527112007141113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,16384,0.8654079967074924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,16384,1.7945910559760199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,1024,0.1854942242304484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,65536,3.395977867974175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,12288,0.5939315689934624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,65536,7.218149820963542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,12288,1.3742133246527777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,10240,0.4676453272501628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,10240,1.142732408311632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,8192,0.37710221608479816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,8192,0.9153315226236979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,512,0.1049484411875407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,128,0.059227552678849965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,7168,0.3341626591152615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,256,0.07255378034379747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,7168,0.8049902386135526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,6144,0.2890728844536675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,6144,0.6931262016296387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,5120,0.24123822318183052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,5120,0.6120808919270834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,16384,2.283760918511285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,4096,0.19919378227657744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,4096,0.4939369095696344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,12288,1.7303529315524633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,3584,0.17336532804701063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,3584,0.4336853292253282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,10240,1.4467510647243922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,3072,0.15095199478997126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,65536,9.165496826171875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,3072,0.3737200101216634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,2560,0.12976711326175266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,2560,0.31589778264363605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,8192,1.1601218117607963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,2048,0.10625600152545506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,2048,0.2540293269687229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,7168,1.0258489184909396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,6144,0.8906275431315104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,1536,0.08453244633144802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,1536,0.19203466839260527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,5120,0.7436586485968696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,1024,0.061703112390306264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,1024,0.13187466727362737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,768,0.04986222253905403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,768,0.09674666987525092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,3584,0.534621344672309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,512,0.03868533174196879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,512,0.06869777705934313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,2560,0.39073954688178164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,256,0.027656000521447923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,4096,0.6029475529988607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,256,0.04740711053212484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,128,0.021866665946112737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,128,0.03706666827201843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,3072,0.4624062114291721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,64,0.019461333751678467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,64,0.03445511062939962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,32,0.02013422217633989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,32,0.033936887979507446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,1536,0.25055021709865993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,2048,0.32202667660183376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,768,0.13670044475131565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,16384,0.5909040239122179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,16384,1.7296595043606227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,65536,2.6160863240559897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,1024,0.1759528848859999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,12288,0.4407920042673747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,65536,6.905264960394965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,12288,1.3029164208306205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,10240,0.371251556608412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,10240,1.090559959411621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,8192,0.30057067341274685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,8192,0.8723920186360677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,512,0.09961244132783677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,7168,0.26770933469136554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,256,0.06922399997711182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,7168,0.7696008682250977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,6144,0.22777512338426378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,6144,0.6626648902893066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,16384,2.193609873453776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,5120,0.19164089361826578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,5120,0.5822426478068033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,12288,1.6592071321275499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,4096,0.1561182207531399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,4096,0.4703057607014974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,10240,1.3982470830281575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,3584,0.1393351157506307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,3584,0.41161688168843585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,65536,8.769046359592014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,8192,1.1163582271999783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,3072,0.12373777230580647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,3072,0.3553368780348036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,2560,0.10300177998012966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,2560,0.2998310989803738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,7168,0.9823830922444662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,2048,0.08540888627370198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,2048,0.24196799596150717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,6144,0.8507173326280383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,1536,0.06724711259206136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,1536,0.18493244383070204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,3584,0.5118693245781792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,1024,0.04940711127387153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,5120,0.711551136440701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,1024,0.12531288464864096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,768,0.0402817792362637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,2560,0.37528088357713485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,768,0.09044888946745132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,512,0.031131555636723835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,512,0.06408444378111097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,4096,0.5781333181593153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,256,0.021949332621362474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,256,0.04404177930619982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,128,0.05571733580695259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,128,0.017224889662530687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,3072,0.4438569015926785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,128,0.03394133183691237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,64,0.014850666125615438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,64,0.03154844376775954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,32,0.01552977826860216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,32,0.03030666708946228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,1536,0.24132355054219565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,2048,0.3088364336225721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,65536,1.9666116502549913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,16384,0.44089778264363605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,16384,1.6585867140028212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,65536,6.6277830335829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,12288,0.33477600415547687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,12288,1.2606995900472004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,256,0.06490311357710096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,10240,0.28195378515455455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,128,0.05287733342912462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,512,0.09453155597050984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,8192,0.22788355085584852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,10240,1.0521440505981445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,8192,0.8422435654534234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,7168,0.19942755169338652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,7168,0.7421422004699707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,768,0.1297475496927897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,6144,0.17449511422051323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,6144,0.6385893291897243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,1024,0.1696088843875461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,5120,0.14660355779859754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,5120,0.5505466461181641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,16384,2.116395526462131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,4096,0.12071822749243842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,4096,0.44554842842949766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,12288,1.591128879123264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,65536,8.424620734320746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,3584,0.1057351099120246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,3584,0.3936737908257379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,3072,0.09408444166183472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,3072,0.3387760056389703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,10240,1.3331208758884006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,2560,0.07886577977074517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,8192,1.07423824734158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,2560,0.2861875428093804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,2048,0.06549066967434354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,2048,0.23112087779574922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,7168,0.9435964160495334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,1536,0.05208355519506666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,1536,0.17671910921732584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,6144,0.8189333279927572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,1024,0.038634667793909706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,1024,0.119832886589898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,5120,0.684241771697998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,768,0.03238133258289761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,768,0.08637866708967422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,3584,0.4920310974121094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,512,0.024801777468787298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,512,0.061347557438744434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,4096,0.5551262431674534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,256,0.017656889226701524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,256,0.04093955622778998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,2560,0.3598951233757867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,128,0.013607111242082385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,128,0.030814223819308813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,3072,0.4248995516035292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,64,0.011857777833938599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,64,0.02764355474048191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,32,0.0122417774465349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,32,0.025851555996470984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,1536,0.23090932104322645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,2048,0.2978106604682075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,65536,1.1652008692423503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,16384,0.294177770614624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,16384,1.5832924313015406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,768,0.12382400035858154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,12288,0.22353778945075142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,65536,6.2874399820963545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,12288,1.1955697801378038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,10240,0.18687910503811309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,10240,1.000314712524414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,1024,0.16176266140407985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,8192,0.1523804399702284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,8192,0.8023004531860352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,256,0.06181866592831082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,7168,0.13381155331929526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,7168,0.7076773113674588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,512,0.09062488873799641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,6144,0.11597600248124863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,128,0.04886133472124735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,6144,0.6097840203179253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,5120,0.09791022539138794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,5120,0.5196888711717393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,16384,2.0268736945258246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,4096,0.07953155703014798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,4096,0.42046133677164715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,12288,1.5271422068277996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,3584,0.07068800263934664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,3584,0.3698977894253201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,65536,8.04567125108507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,3072,0.0636684430970086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,3072,0.3212488757239448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,10240,1.279107517666287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,2560,0.05273422267701891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,2560,0.2695315678914388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,8192,1.0345911449856229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,2048,0.04454844527774387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,2048,0.2181253300772773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,7168,0.9054230584038628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,1536,0.03495733274353875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,6144,0.7792737748887805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,1536,0.1668213340971205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,1024,0.025563556287023757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,1024,0.11276266309950088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,5120,0.6576364305284288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,768,0.02087822225358751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,768,0.08120799726910062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,3584,0.4706186718410916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,512,0.016179554992251925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,512,0.05605066484875149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,2560,0.343486229578654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,256,0.011336888704035016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,256,0.03392533461252848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,4096,0.5315697987874349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,128,0.008778666456540426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,128,0.025384889708624944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,3072,0.40818222363789874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,1536,0.21977689531114367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,64,0.007639110916190677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,64,0.022632888621754114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,32,0.008375999828179678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,32,0.021457778082953557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,768,0.11664444870418972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,2048,0.2858915593889025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,65536,1.028029335869683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,16384,0.22678399085998535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,16384,1.5812035666571722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,12288,0.1705848905775282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,12288,1.189503139919705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,1024,0.1535537772708469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,65536,6.242489708794488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,10240,0.14387555917104086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,10240,0.9960266749064127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,8192,0.1168702178531223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,8192,0.8001466857062446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,512,0.085372441344791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,256,0.054210666153166026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,7168,0.10309600167804295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,7168,0.7022124396430122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,6144,0.08857155508465236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,6144,0.6063013076782227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,128,0.041856888267729014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,5120,0.07545422183142768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,5120,0.5051377614339193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,16384,1.9834889305962458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,4096,0.0614337788687812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,4096,0.4070853392283122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,12288,1.4932995902167425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,3584,0.05504088931613498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,10240,1.248644404941135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,3584,0.3589182164933946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,3072,0.048173334863450795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,3072,0.31020887692769367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,65536,7.830775790744358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,2560,0.04199022385809156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,2560,0.2612888813018799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,8192,1.0036630630493164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,2048,0.03426133261786567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,2048,0.21188533306121826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,7168,0.8830657535129123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,6144,0.7621253331502279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,1536,0.027476444840431213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,1536,0.16236444314320883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,1024,0.02015466656949785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,1024,0.11177155706617568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,5120,0.6398782200283474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,768,0.016232000456915963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,768,0.08093688885370891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,3584,0.459645324283176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,512,0.012887111140622033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,512,0.05419111251831055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,2560,0.33838756879170734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,256,0.009144888983832465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,256,0.03246666656600104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,4096,0.5209031105041504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,128,0.0073520003093613525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,128,0.023927110764715407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,3072,0.39919021394517684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,64,0.006484444356626934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,64,0.021282666259341772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,32,0.006801777829726537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,32,0.01995733380317688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,768,0.11420267158084446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,2048,0.27677422099643284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,65536,0.6056657897101508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,16384,0.16014311048719618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,1536,0.213137772348192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,16384,1.4958062701755102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,12288,0.11537333329518636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,12288,1.1302684148152669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,65536,5.953062269422744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,10240,0.1030399998029073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,10240,0.9494720035129123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,1024,0.14885332849290636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,8192,0.07903466622034709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,8192,0.7593226432800293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,128,0.03976800044377645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,7168,0.07026311424043444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,7168,0.6667164696587456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,256,0.05283911029497782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,6144,0.062071111467149526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,6144,0.5763529141743978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,512,0.0834124419424269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,5120,0.052314665582444936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,5120,0.4879893196953668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,16384,1.9400914510091145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,4096,0.04344888859324985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,4096,0.3946151203579373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,12288,1.4621707068549261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,3584,0.039226667748557195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,3584,0.34781776534186465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,10240,1.223137749565972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,3072,0.03435644507408142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,3072,0.3014640013376872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,8192,0.9861173629760742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,2560,0.029820445511076186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,65536,7.6878568861219625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,2560,0.25358133845859104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,2048,0.02479733361138238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,2048,0.20554044511583117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,7168,0.8677635722690158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,1536,0.019143111175960965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,1536,0.1570168866051568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,6144,0.7486737569173177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,1024,0.014320888453059726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,1024,0.10806400246090359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,5120,0.6295733451843262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,768,0.012448000411192576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,768,0.07671644290288289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,3584,0.4506568908691406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,512,0.009676444033781687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,512,0.05149866806136238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,4096,0.5108213424682617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,256,0.007425777614116669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,256,0.0302471121152242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,2560,0.3291884528266059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,128,0.006177777631415262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,128,0.022325333621766832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,3072,0.39078132311503094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,64,0.005588444570700328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,64,0.01960088809331258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,32,0.005809777726729711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,32,0.018375999397701688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,65536,0.41250401073031956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,2048,0.26957154273986816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,1536,0.20796799659729004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,16384,0.10470310846964519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,65536,5.885963439941406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,16384,1.48815737830268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,12288,0.08068355586793688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,12288,1.1208213170369465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,768,0.11132000552283393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,10240,0.06875022252400716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,10240,0.9386079576280383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,1024,0.1444586647881402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,8192,0.055424888928731285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,256,0.05158844590187073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,8192,0.7549395561218262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,7168,0.049662222464879356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,512,0.08174666431215075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,7168,0.6640328831142849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,6144,0.04366488920317756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,6144,0.571905771891276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,16384,1.9189075893825953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,5120,0.03746577766206529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,5120,0.47392087512546116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,65536,7.5766389634874125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,4096,0.03475911087459988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,12288,1.442616886562771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,4096,0.3826879925198025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,3584,0.02795377704832289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,3584,0.3372026549445258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,10240,1.2076702117919922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,3072,0.024726221958796184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,3072,0.2916782167222765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,8192,0.9693235821194119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,2560,0.02160533269246419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,7168,0.8534595701429578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,2560,0.24578399128384062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,2048,0.013813333378897773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,2048,0.19945777787102592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,6144,0.7365475760565864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,1536,0.010805333654085795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,1536,0.15259377161661783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,5120,0.6206986639234755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,1024,0.00848444468445248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,1024,0.10463822550243801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,3584,0.44194311565823025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,768,0.007306666837798224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,4096,0.5021048651801215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,768,0.07415911224153307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,512,0.006184000107977126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,512,0.04957155717743767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,2560,0.32317688730027944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,3072,0.3836266729566786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,256,0.005150222116046482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,256,0.028760890165964764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,128,0.004459555364317364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,128,0.02100177771515316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,2048,0.2637244330512153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,64,0.004091555459631814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,64,0.01827555563714769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,32,0.004318222403526306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,32,0.0170151111152437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,65536,0.4168719980451796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,1536,0.20380177762773302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,65536,5.887392679850261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,16384,0.1253653367360433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,16384,1.488077375623915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,128,0.03896177808443705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,12288,0.07486755318111844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,12288,1.122137811448839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,512,0.08003999789555867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,10240,0.0631635586420695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,1024,0.14196711116366917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,768,0.10815644264221191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,10240,0.9381643931070963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,8192,0.05059822069274055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,8192,0.7545288933648003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,256,0.050056000550587974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,7168,0.04496622085571289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,128,0.03712000118361579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,7168,0.6636551221211752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,6144,0.04011377692222595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,6144,0.5721413294474283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,5120,0.03384088807635837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,5120,0.4722088707817926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,10240,1.1989911397298176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,4096,0.028047998746236164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,4096,0.38077333238389754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,16384,1.8987483978271484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,3584,0.026034666432274714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,3584,0.3365644348992242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,12288,1.4330889383951824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,3072,0.022535110513369244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,3072,0.2911449008517795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,8192,0.9654479556613498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,2560,0.0177839994430542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,2560,0.24532177713182238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,5120,0.6139538023206923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,2048,0.010016889207892949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,7168,0.8478453424241809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,2048,0.19969156053331164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,1536,0.007825777762466008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,1536,0.1526542239718967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,6144,0.7308150927225748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,1024,0.006068444500366847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,1024,0.10342488686243693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,65536,7.514376322428386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,768,0.005496888938877318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,768,0.07419466972351074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,4096,0.4969751040140788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,3584,0.43815022044711643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,512,0.004912000149488449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,512,0.04958933260705736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,256,0.004107555581463708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,256,0.028919110695521038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,2560,0.32048355208502877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,128,0.003682666768630346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,3072,0.37932266129387754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,128,0.02070399953259362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,64,0.003338666632771492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,64,0.01771377854877048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,32,0.0037093332244290244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,2048,0.2614719867706299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,32,0.016384000579516094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,65536,0.3860729005601671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,1536,0.20077778233422172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,16384,0.09432799948586358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,12288,0.07195111115773518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,12288,1.1184977425469291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,16384,1.4897520277235243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,10240,0.06031022469202677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,8192,0.049478222926457725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,10240,0.9385555055406359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,8192,0.7542791366577148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,7168,0.04359822140799629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,6144,0.03839911023775736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,1024,0.13942044311099583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,7168,0.6624222331576878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,5120,0.03251466817326016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,6144,0.5721617804633247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,4096,0.026962666047943964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,3584,0.024002666274706524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,5120,0.47288889355129665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,4096,0.38344534238179523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,3072,0.021320000290870667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,3584,0.3371680047776964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,2560,0.01627377834584978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,2048,0.009564444422721863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,3072,0.2924862172868517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,2560,0.24616177876790366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,1536,0.008172444171375698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,2048,0.20006310939788818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,1024,0.0064284445510970215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,768,0.005557333429654439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,1536,0.15265599886576334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,1024,0.10367466343773736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,512,0.0046737777690092725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,768,0.07407022184795804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,256,0.0038844446341196695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,512,0.0498560004764133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,128,0.0034844444857703317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,256,0.029085334804322984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,64,0.0033564445459180405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,128,0.020738666256268818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,65536,5.888695187038845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,32,0.003436444534195794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,64,0.017792888813548617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,32,0.016301333904266357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,16384,0.09335733122295803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,12288,0.07157688670688205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,768,0.10681244399812485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,65536,0.37822845247056747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,10240,0.05999733342064751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,8192,0.04903022117084927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,12288,1.1222249137030709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,16384,1.4883217281765406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,10240,0.937031110127767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,7168,0.04435111085573832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,6144,0.03806489043765598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,7168,0.6626337899102105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,8192,0.7558355331420898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,5120,0.0321760012043847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,6144,0.5714444584316677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,4096,0.026552889082166884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,3584,0.023524444964196947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,5120,0.4743279880947537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,4096,0.3831377824147542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,3072,0.0207315550910102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,3584,0.33778222401936847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,2560,0.015412444869677225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,2048,0.008489777644475302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,3072,0.29190754890441895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,2560,0.24589334593878853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,1536,0.007064888874689738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,2048,0.20048533545600045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,1024,0.005534222142563925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,768,0.004996444616052839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,1536,0.15306933720906576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,1024,0.10397955444124009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,512,0.004126222183307012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,768,0.07363822062810262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,256,0.003679111186001036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,512,0.04923466510242886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,128,0.003343111111058129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,256,0.029069334268569946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,64,0.003072888900836309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,128,0.02086399992307027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,32,0.0031617778456873367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,64,0.017641777793566387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,32,0.016358221570650738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,65536,5.883780585394965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,256,0.05037422312630547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,512,0.0787902209493849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,12288,6.795353783501519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,16384,8.665475633409288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,12288,5.560227711995442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,16384,7.384453667534722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,10240,5.561097039116754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,10240,6.680928972032334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,12288,7.991634792751736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,10240,4.621228535970052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,16384,10.628362019856771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,8192,3.702365451388889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,7168,3.22999636332194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,8192,3.7403733995225696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,7168,3.400339550442166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,8192,5.351715511745876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,6144,2.7914649115668406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,128,0.03624533282385932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,6144,4.0116221110026045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,7168,4.6847351921929254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,5120,2.309549331665039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,6144,2.860369788275825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,5120,3.4021263122558594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,4096,1.8450497521294489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,5120,2.4021502600775824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,4096,1.8723013136121962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,3584,1.6212435828314886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,3072,1.3913164138793945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,4096,2.728025860256619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,3584,1.6830835342407227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,2560,1.1586951149834526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,3584,2.4193431006537542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,3072,1.400895118713379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,3072,2.1207154591878257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,2560,1.178144031100803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,2048,0.9216595755683051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,2560,1.7792445288764105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,1536,0.6848728921678332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,2048,0.9408915837605795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,1536,0.725652429792616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,1024,0.45710844463772243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,2048,1.4334808985392253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,1024,0.7034631305270724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,768,0.3493671152326796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,1536,1.001871109008789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,1024,0.5159937540690104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,512,0.28585865762498647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,768,0.4269057909647624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,256,0.273726224899292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,256,0.28998398780822754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,512,0.4117102093166775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,128,0.2248471048143175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,256,0.2705191241370307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,512,0.31650932629903156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,768,0.5462409125434028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,64,0.20760444800059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,32,0.2036986615922716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,64,0.22764976819356283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,128,0.25815023316277397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,32,0.2151262230343289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,128,0.260463105307685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,12288,1.8139875200059679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,16384,2.3675645192464194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,16384,2.406277338663737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,12288,2.0800374348958335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,12288,1.8906870947943792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,10240,1.5496977700127497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,16384,2.7828409406873913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,10240,1.7339884440104167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,10240,1.5343590842352972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,8192,1.174736870659722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,8192,1.401603592766656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,8192,1.2264524035983615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,7168,1.0468720330132377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,7168,1.0792995029025607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,7168,1.2249999576144748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,6144,0.9107333289252387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,6144,1.0794764624701607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,5120,0.7569973203870984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,6144,0.919832017686632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,5120,0.8553742302788628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,5120,0.7797857920328776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,4096,0.6535458034939235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,4096,0.6060115496317545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,4096,0.6311795446607802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,3584,0.5339822239345974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,3584,0.592201762729221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,3584,0.5642959806654189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,3072,0.47962310579087997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,3072,0.4593591160244412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,2560,0.3832968870798747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,3072,0.48308176464504665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,2560,0.4063902166154649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,2560,0.40741509861416286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,2048,0.3069840007358127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,1536,0.22408000628153482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,2048,0.3417244487338596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,2048,0.33395200305514866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,1536,0.2663715680440267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,1536,0.2568417655097114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,1024,0.15129689375559488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,768,0.14128711488511828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,768,0.1186479992336697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,1024,0.19229777654012045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,1024,0.1850115590625339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,512,0.10659999979866876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,512,0.0941271119647556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,256,0.06696889135572645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,768,0.1574017736646864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,512,0.12063911226060654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,256,0.07776444488101535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,128,0.05935021903779772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,256,0.09245066510306464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,64,0.05548355645603604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,128,0.06990844673580594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,32,0.05591111050711738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,128,0.08724711338678996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,64,0.07015466690063477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,32,0.06830488973193698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,65536,13.154871622721354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,65536,9.915345933702257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,16384,1.9311254289415147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,16384,2.0336231655544705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,12288,1.490989367167155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,16384,2.038512971666124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,12288,1.481737772623698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,12288,1.5272817611694336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,10240,1.1604319678412545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,65536,10.96012963189019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,10240,1.2544977400037978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,10240,1.2885715696546767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,8192,0.9134604136149088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,8192,0.9751173655192057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,8192,1.0434533225165474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,7168,0.8799200057983398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,7168,0.7768719991048177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,7168,0.8996746275160047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,6144,0.7582142088148328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,6144,0.7017671267191569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,5120,0.6292844348483616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,6144,0.7966933250427246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,65536,9.74746788872613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,5120,0.659804450141059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,4096,0.45394309361775714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,5120,0.5670346683926052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,4096,0.50528531604343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,3584,0.44640975528293186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,4096,0.537102222442627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,3584,0.4012426535288493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,3584,0.48078134324815536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,3072,0.38306310441758895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,2560,0.31983910666571724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,3072,0.4115751054551866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,3072,0.3474915557437473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,2048,0.25527999136182994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,2560,0.3577866554260254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,2048,0.2910044458177355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,2560,0.29831822713216144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,1536,0.18622667259640166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,2048,0.23945244153340658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,1024,0.1345635520087348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,1536,0.22674666510687935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,1024,0.12690399752722845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,768,0.10696444246504043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,1024,0.1637279987335205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,768,0.10029777553346421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,1536,0.18551111221313477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,512,0.08145955536100599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,768,0.13668621910942927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,512,0.07916711436377631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,256,0.05191911260286967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,512,0.10544444455040826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,256,0.06439022223154704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,128,0.04667822188801236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,256,0.08050400018692017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,64,0.04263910982343885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,32,0.04363822274737888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,128,0.06263644165462919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,128,0.07440800136990018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,64,0.057550224992964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,32,0.05660444498062134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,65536,8.42246331108941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,16384,1.7223457760281031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,16384,1.8584728240966797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,16384,1.7130444844563801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,12288,1.3318701850043404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,12288,1.3213457531399198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,12288,1.4001582463582356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,10240,1.1194417741563585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,10240,1.096655951605903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,65536,7.8920398288302955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,8192,0.8784471087985568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,10240,1.16321775648329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,8192,0.9419067170884875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,8192,0.7614453103807238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,7168,0.6981324619717069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,7168,0.7811137835184733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,7168,0.8264524671766492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,6144,0.678017775217692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,6144,0.5949991014268663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,5120,0.4912533230251736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,6144,0.7162906858656141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,5120,0.5655306710137261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,4096,0.4538257916768392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,5120,0.5984533098008897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,4096,0.49445602628919816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,4096,0.3866275681389703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,3072,0.34377598762512207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,3584,0.3484951125250922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,3584,0.39823023478190106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,65536,7.5414479573567705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,3584,0.43171734280056423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,3072,0.376310216055976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,2560,0.28801245159573025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,3072,0.3040613333384196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,2048,0.23052000999450684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,1536,0.16428532865312365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,65536,8.37219492594401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,2048,0.26362488004896373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,2560,0.3264799912770589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,2560,0.2533155547247993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,768,0.09489066733254327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,1024,0.11368710464901394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,1024,0.11425333552890354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,1536,0.1675128936767578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,1024,0.15089600616031223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,2048,0.2068799999025133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,1536,0.20439288351270887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,768,0.09043911430570815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,512,0.071706665886773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,512,0.07181333170996772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,256,0.04912088976966011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,768,0.1210284498002794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,128,0.03971911138958401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,256,0.05815111266242134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,512,0.09828977452384101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,256,0.0750817788971795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,64,0.035972442891862653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,128,0.050332443581687085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,128,0.06689333253436618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,32,0.037017779217825994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,64,0.05096088846524557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,32,0.05027733246485392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,16384,1.516084459092882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,12288,0.957148445977105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,16384,1.6566818025377061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,65536,6.752824147542317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,16384,1.3148657480875652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,10240,0.9826462003919813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,12288,1.1692746480305989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,12288,1.2556986278957791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,10240,0.7477911313374838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,8192,0.7750088903639051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,8192,0.849731551276313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,8192,0.626164436340332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,10240,1.0535590913560655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,7168,0.5235190921359593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,7168,0.6857137680053711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,6144,0.5984942118326823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,7168,0.7485537528991699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,6144,0.45256180233425564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,6144,0.640734248691135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,5120,0.49785598119099933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,4096,0.3997982343037923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,5120,0.5471199883355035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,4096,0.30551287862989635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,5120,0.37749867969089085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,3584,0.35140265358818906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,3584,0.3959893385569255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,4096,0.44368799527486164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,3584,0.2766488922966851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,3072,0.304367118411594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,3072,0.23447023497687447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,3072,0.34107110235426163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,2560,0.25496000713772243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,2560,0.19894933700561523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,2560,0.29350754949781627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,2048,0.20420977804395887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,1536,0.12669777870178223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,2048,0.1640311082204183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,2048,0.239153782526652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,1536,0.14874311288197836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,65536,6.825985802544488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,1024,0.0916702217525906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,768,0.07536977529525757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,1024,0.10049421919716729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,1536,0.1862177716361152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,512,0.056417776478661426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,65536,6.753811730278863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,768,0.07981866598129272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,1024,0.13442577256096735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,512,0.06404533651140001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,768,0.10992444223827785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,256,0.03695111142264472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,512,0.08628266387515598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,128,0.032285332679748535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,256,0.05429333448410034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,256,0.06534755229949951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,128,0.04355555441644457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,64,0.02932000160217285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,32,0.030208001534144085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,128,0.06045955419540405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,64,0.04454400142033895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,32,0.04343466626273262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,16384,1.4183182186550563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,12288,0.860761801401774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,16384,1.5942409303453233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,16384,1.1500346925523546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,12288,1.0834870868259006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,12288,1.192554685804579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,10240,0.6931129031711154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,65536,5.158048841688368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,10240,0.9214942720201281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,10240,0.9899475309583875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,8192,0.7176951302422417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,8192,0.5331981976826986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,7168,0.6417386796739366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,8192,0.8037608994377984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,7168,0.460584905412462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,7168,0.7140355639987521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,6144,0.3975333372751872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,6144,0.5573582119411892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,5120,0.46350844701131183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,6144,0.6130951245625814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,5120,0.33166665501064724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,4096,0.3741226726108127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,5120,0.5191004541185167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,4096,0.26819732454088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,4096,0.42336355315314395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,3584,0.32777243190341526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,3072,0.28475112385219997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,65536,6.185007307264541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,3584,0.3755226665072971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,3072,0.32583644655015737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,3584,0.23771111170450845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,3072,0.20767911275227866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,2560,0.23814755015903047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,2560,0.2787911097208659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,2048,0.14486666520436606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,2048,0.19107377529144287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,1536,0.1105804443359375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,2560,0.1766995588938395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,2048,0.22652800877888998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,1536,0.1389040019777086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,1024,0.08218844731648763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,1536,0.17728266451093885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,1024,0.09436177545123631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,768,0.06539022260242038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,768,0.07629777987798055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,1024,0.12882222069634333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,512,0.05058311091528999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,512,0.05959022045135498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,768,0.10579999950197007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,256,0.03356000118785434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,512,0.08257777823342217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,256,0.04712177647484673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,256,0.06372444497214423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,128,0.02909066610866123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,128,0.040992889139387347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,32,0.02723377777470483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,64,0.02689244515366024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,128,0.05747822258207533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,64,0.04099466734462314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,32,0.040213333235846624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,65536,6.300359937879775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,16384,1.3204275767008464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,16384,0.9993368784586588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,16384,1.4847289191351996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,12288,0.6738995446099175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,12288,1.0165341695149739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,12288,1.1148897806803386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,65536,4.5152363247341585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,10240,0.8482800059848361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,10240,0.5590551164415147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,10240,0.9347759882609049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,8192,0.6677235497368706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,8192,0.45053778754340273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,8192,0.7550462086995443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,7168,0.5972026718987359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,7168,0.39602043893602157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,7168,0.6682524681091309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,6144,0.34166844685872394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,6144,0.5176951090494791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,5120,0.43001333872477215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,65536,5.4130605061848955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,6144,0.5731689135233561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,5120,0.48923465940687394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,4096,0.3487199942270915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,5120,0.2855902247958713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,4096,0.2333786752488878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,4096,0.39662665790981716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,3584,0.30637065569559735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,3584,0.353911108440823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,3072,0.2637831105126275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,2560,0.154621336195204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,3072,0.3046355512407091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,3584,0.20863821771409777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,2048,0.12705600261688232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,2560,0.2225715584225125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,3072,0.17872356043921578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,2560,0.2609395451015896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,1536,0.09765422344207764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,2048,0.1772977775997586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,1536,0.12884710894690618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,1024,0.07020799981223212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,2048,0.21415111753675672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,1536,0.16558310720655653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,1024,0.08801155620151097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,768,0.058005332946777344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,65536,5.823040008544922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,768,0.07086755832036336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,512,0.044028444422615894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,1024,0.11955555280049641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,512,0.056023114257388644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,768,0.10016800297631158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,256,0.032299555010265775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,256,0.043649779425727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,512,0.07916355795330472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,128,0.025779555241266888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,256,0.060621334446801074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,128,0.03726933399836222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,64,0.024339555038346186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,128,0.054126222928365074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,32,0.024519999821980793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,64,0.03758577836884393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,32,0.03708800011210971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,12288,0.5759030977884928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,16384,1.2218222088283963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,16384,1.3914960225423176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,16384,0.8442702293395996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,65536,3.8628090752495656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,12288,0.9304951561821831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,10240,0.4762079980638292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,12288,1.0415289137098525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,10240,0.7869111167060004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,8192,0.37850220998128253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,8192,0.620441754659017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,10240,0.8796026441786023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,8192,0.711027569240994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,7168,0.5473315450880263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,7168,0.3357688850826687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,6144,0.4767395655314128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,7168,0.6275102297465006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,6144,0.2919520007239448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,6144,0.5437111324734157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,5120,0.39726487795511883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,5120,0.2431137826707628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,4096,0.3210293451944987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,5120,0.4571350945366754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,4096,0.3765546745724148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,4096,0.20016088750627306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,3584,0.28274310959710014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,3584,0.33550222714742023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,3072,0.15417244699266222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,3584,0.17538489235772026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,3072,0.24372445212470162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,2560,0.13274755742814806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,65536,5.00106684366862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,3072,0.29081866476270885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,2560,0.20495022667778862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,2048,0.10707377725177342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,1536,0.08440622356202866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,1536,0.11936888429853652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,2048,0.16443999608357748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,2560,0.24598577287462023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,1024,0.06173155705134074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,1536,0.15454310841030544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,2048,0.19992710484398735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,768,0.05022311210632324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,1024,0.08156533373726739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,768,0.06632266441980998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,1024,0.11321777767605251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,512,0.038677334785461426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,256,0.02740977704524994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,768,0.09465510977639092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,512,0.05224177903599209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,512,0.07345689005321927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,128,0.022431999444961548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,256,0.03953866826163398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,256,0.057447108957502574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,128,0.03408266769515143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,64,0.01975733372900221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,128,0.05001688996950785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,32,0.020671111014154222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,64,0.03504088852140639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,32,0.03369777732425266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,65536,5.466796451144749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,16384,1.1085733837551541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,16384,0.6162195735507542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,16384,1.2938808865017362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,12288,0.45467021730211044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,12288,0.8460328843858508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,65536,3.3834302690294056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,12288,0.9810098012288412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,10240,0.37482134501139325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,10240,0.7093377643161349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,8192,0.5604826609293619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,8192,0.29962489340040416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,8192,0.6611430909898546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,10240,0.8196409013536242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,7168,0.4930097791883681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,6144,0.4304204516940647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,7168,0.26366400718688965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,7168,0.5859368642171224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,6144,0.22782489988538954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,6144,0.5110053486294216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,5120,0.35608532693650985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,4096,0.2892826663123237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,5120,0.4331839879353841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,5120,0.19217956066131592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,4096,0.1561431090037028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,3584,0.13964533805847168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,4096,0.3507555590735541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,3072,0.12042400572035049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,3584,0.2500106758541531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,3072,0.21541778246561685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,3584,0.31034310658772785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,2560,0.10284088717566596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,2560,0.18136444356706408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,3072,0.26919733153449166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,2048,0.08591910865571763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,2048,0.14331644111209446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,1536,0.06662399901284112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,2560,0.23315644264221191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,1536,0.10263911220762466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,2048,0.1905777719285753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,1024,0.04878133204248217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,1024,0.07126844591564603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,768,0.04081333345837063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,1536,0.1450631088680691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,65536,4.443330552842882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,768,0.058803558349609375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,512,0.0314062237739563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,1024,0.1060568888982137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,512,0.04698933495415581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,768,0.08886844582027859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,256,0.021611554755104914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,256,0.03545777665244208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,128,0.016470222009552848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,512,0.06866311364703707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,64,0.014699555105633207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,128,0.03033422099219428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,256,0.05394666724734836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,128,0.04656444324387444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,32,0.015477332803938123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,32,0.02975911233160231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,64,0.030060443613264296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,65536,2.6056319342719183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,65536,5.116730584038629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,16384,1.0599368413289387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,12288,0.4195653332604303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,12288,0.8115190929836698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,16384,0.5551253424750434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,10240,0.35922222667270237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,10240,0.682371563381619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,8192,0.5381857554117838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,8192,0.2849208778805203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,7168,0.2503280109829373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,7168,0.4718426598442926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,65536,4.222721099853516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,65536,2.5831288231743708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,6144,0.2161031166712443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,6144,0.41097333696153426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,5120,0.18377955754597983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,5120,0.34150489171346027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,16384,1.2897368537055123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,4096,0.14804711606767443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,4096,0.27809866269429523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,12288,0.9886052873399523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,3584,0.13133600023057726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,3584,0.24230310651991102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,10240,0.8290213478936089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,3072,0.1153528822792901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,8192,0.6711484591166178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,3072,0.20906044377221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,2560,0.09841066598892212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,2560,0.17638756169213188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,7168,0.5910462273491753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,2048,0.08103733592563205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,2048,0.13902400599585638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,65536,5.140795389811198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,1536,0.06595110893249512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,1536,0.10073422061072455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,6144,0.5121688842773438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,1024,0.04807110958629184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,1024,0.06962222523159452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,5120,0.4339333375295003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,768,0.03896533449490865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,768,0.05681777662701077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,3584,0.3147466712527805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,512,0.02910755409134759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,512,0.04550488789876302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,2560,0.2331848939259847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,256,0.020040000478426617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,4096,0.3543875482347276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,256,0.03365866674317254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,128,0.015072888798183866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,128,0.02880177895228068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,3072,0.2731422318352593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,64,0.014296889305114746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,64,0.02805866797765096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,32,0.015285332997639975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,32,0.0277786652247111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,1536,0.14689511722988552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,2048,0.190774228837755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,768,0.08969689077801174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,1024,0.10793866713841756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,65536,4.0115174187554254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,16384,0.4422115484873454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,16384,1.00908268822564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,65536,2.0288933648003473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,16384,1.202954716152615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,12288,0.3365360101064046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,65536,4.76669438680013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,10240,0.6472053527832031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,12288,0.7704364458719889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,10240,0.2830631203121609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,10240,0.7705493503146701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,12288,0.9067057503594292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,8192,0.5122266875372993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,8192,0.22664088673061797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,7168,0.45343377855088973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,8192,0.6240835719638401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,7168,0.20161333349015978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,5120,0.14744622177547878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,7168,0.5451466772291396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,6144,0.39437956280178493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,6144,0.17295377784305146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,6144,0.4787439770168728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,4096,0.12220533688863118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,5120,0.32735999425252277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,4096,0.2647644413842095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,3584,0.10886400275760227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,5120,0.39760979016621906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,3072,0.09480533334943984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,3584,0.23342490196228027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,4096,0.33048976792229545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,3584,0.29106399748060435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,3072,0.20071821742587617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,2560,0.0790826678276062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,2560,0.1685404380162557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,2048,0.06655200322469075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,3072,0.2548977798885769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,2048,0.13446044921875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,1536,0.052685333622826465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,2560,0.21600800090365938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,1536,0.09676711426840888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,1024,0.03906844390763177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,2048,0.17808088991377088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,1024,0.06663111183378431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,1536,0.1346328920788235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,768,0.03198933270242479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,512,0.024433778391944036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,768,0.05512977639834086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,1024,0.09988266891903347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,512,0.04361600014898512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,768,0.08248266908857557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,256,0.018030222919252183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,512,0.06573244598176745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,256,0.0321724447939131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,128,0.013800000150998434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,128,0.02721422248416477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,64,0.012038222617573209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,256,0.050386667251586914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,32,0.01254755589697096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,64,0.024664888779322307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,128,0.041783110962973706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,32,0.024342222346199885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,128,0.04652355445755852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,256,0.054583999845716685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,16384,0.3711671034495036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,512,0.07119377454121907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,16384,0.9630275302463107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,12288,0.7327822049458822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,12288,0.2800248993767632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,10240,0.23479111989339194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,10240,0.6154017978244358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,8192,0.4874720043606228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,8192,0.18962489234076607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,7168,0.16701689031389025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,65536,3.8036961025661893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,65536,1.719767994350857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,7168,0.4289831055535211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,6144,0.1694177786509196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,6144,0.37243111928304035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,5120,0.12171733379364014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,5120,0.31131911277770996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,16384,1.2097671296861436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,4096,0.09977689054277207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,12288,0.921478271484375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,4096,0.2523555490705702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,10240,0.7689075469970703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,3584,0.08779999944898818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,3584,0.22137778335147432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,3072,0.07677333222495185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,3072,0.19201421737670898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,8192,0.6270373132493761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,2560,0.06636089086532593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,2560,0.16135377354092068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,7168,0.5536346435546875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,2048,0.054584887292650014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,2048,0.1274497773912218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,65536,4.7462035285101996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,1536,0.044439110491010875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,6144,0.4805315335591634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,1536,0.09278311332066853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,1024,0.0317555566628774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,1024,0.06455200248294406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,5120,0.4025368955400255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,768,0.02567200031545427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,768,0.05217777689297994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,3584,0.29150933689541286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,512,0.01977066695690155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,512,0.041371554136276245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,2560,0.21678933832380506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,256,0.013822221921549903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,256,0.0292524430486891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,1536,0.13538132773505315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,128,0.010479999913109673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,4096,0.32969511879814994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,128,0.023564444647894964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,64,0.009366222553782994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,64,0.022836445106400385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,768,0.08369866344663833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,32,0.009415999882751042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,3072,0.2541368802388509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,32,0.022469333476490442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,2048,0.17729243967268202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,16384,0.29633598857455784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,1024,0.10124266809887356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,16384,0.9102000130547417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,65536,1.2905022303263347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,65536,3.6142294141981335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,12288,0.22237245241800943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,12288,0.6955724292331271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,16384,1.1179955800374348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,10240,0.5837866465250651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,10240,0.18666577339172363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,12288,0.841975106133355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,65536,4.391505771213107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,8192,0.4645102288987901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,7168,0.13433867030673557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,8192,0.15079466501871744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,10240,0.7131537861294217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,6144,0.11490133073594834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,8192,0.5756159888373481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,7168,0.4087839921315511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,5120,0.09816977712843154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,6144,0.3546133306291368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,7168,0.5096773571438259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,4096,0.081149333053165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,5120,0.29688623216417104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,6144,0.43774043189154727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,4096,0.24070578151279023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,3584,0.07035199801127116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,5120,0.37102577421400285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,3072,0.0616204473707411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,3584,0.21146044466230604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,4096,0.30655378765530056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,2560,0.052806221776538424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,3072,0.18269244829813638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,3584,0.2714444531334771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,2560,0.15440977944268122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,2048,0.04432000054253472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,3072,0.2359680069817437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,2048,0.12252267201741536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,2560,0.19923822085062662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,1536,0.03450222147835626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,2048,0.16441688272688124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,1024,0.02516088883082072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,1536,0.08766666385862563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,1536,0.12330488363901775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,768,0.020777778493033517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,1024,0.06081689066357083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,1024,0.08960622549057007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,512,0.01626488897535536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,768,0.0503715541627672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,256,0.011382222175598145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,768,0.07684622208277385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,512,0.036935110886891685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,512,0.05881422095828586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,256,0.026530666483773127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,128,0.008676444490750631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,256,0.04223733478122287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,64,0.007605333295133378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,128,0.022114665971861944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,32,0.008075555165608725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,64,0.021560000048743352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,128,0.03612444467014737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,32,0.020925333102544148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,16384,0.22487200631035698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,16384,0.8658506605360242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,65536,1.0158791012234158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,12288,0.1732542249891493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,12288,0.6620364189147949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,10240,0.14305955833858913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,65536,3.416532516479492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,256,0.04911466770701938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,10240,0.5532844331529405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,8192,0.11740711000230576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,8192,0.4424568812052409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,7168,0.10194666518105401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,128,0.04039555456903245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,512,0.066484444671207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,7168,0.38975734180874294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,6144,0.08846311436759101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,6144,0.3368524445427789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,5120,0.07490577962663439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,5120,0.28274133470323354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,12288,0.8102266523573133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,4096,0.061262223455641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,4096,0.2286773257785373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,16384,1.0731208589341905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,3584,0.05544444587495592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,3584,0.20213954978519014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,10240,0.6820124520195855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,65536,4.214113023546007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,3072,0.04865866568353441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,8192,0.5496106677585179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,2560,0.04128177629576789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,3072,0.17454666561550566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,2560,0.1469671063952976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,2048,0.03383288780848185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,2048,0.1191377772225274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,7168,0.48423200183444554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,6144,0.4205653402540419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,1536,0.027115555273161993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,1536,0.0854533314704895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,1024,0.019928889142142404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,1024,0.05735377470652262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,5120,0.3570728831821018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,768,0.016521778371598985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,768,0.04611022273699442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,768,0.07240177525414361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,512,0.012757333616415659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,512,0.03529955612288581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,3584,0.2598408857981364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,256,0.009167999857001835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,256,0.02479644450876448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,2560,0.19205689430236816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,128,0.007375111182530721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,128,0.02088533341884613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,1536,0.12046221892038982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,64,0.006425777657164468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,64,0.01983644399378035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,32,0.006709333509206772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,32,0.01956888867749108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,4096,0.2922293345133464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,3072,0.22488533125983345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,2048,0.15598400433858237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,65536,0.6265217992994521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,16384,0.15030845006306967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,1024,0.08742488755120172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,16384,0.821524461110433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,12288,0.11327555444505479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,16384,1.025718265109592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,12288,0.6228826840718588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,10240,0.09528266721301609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,65536,3.215789371066623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,12288,0.7774826685587565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,8192,0.07754044400321113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,10240,0.522781319088406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,10240,0.6519209014044868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,7168,0.06873510943518744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,8192,0.41926222377353245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,8192,0.5303831100463867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,65536,4.034307691786024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,7168,0.37059821022881406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,6144,0.05954933166503906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,7168,0.46805334091186523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,5120,0.05066666669315762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,6144,0.32009956571790904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,6144,0.40615200996398926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,5120,0.2685440116458469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,4096,0.0424488882223765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,5120,0.3422275649176703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,3584,0.038006222910351224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,4096,0.21937244468265107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,3072,0.032778667079077825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,3584,0.19249155786302355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,4096,0.281384891933865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,3584,0.2496408886379666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,2560,0.02796088986926609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,3072,0.16773688793182373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,2048,0.023436443673239812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,3072,0.2158719963497586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,2560,0.14068799548678929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,2560,0.183643552992079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,1536,0.019348444210158456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,2048,0.11445955435434978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,1536,0.08133866389592488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,2048,0.14956088860829672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,1024,0.01404355631934272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,1536,0.11514933904012044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,1024,0.05465777715047201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,768,0.011917333636018964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,1024,0.08433244625727336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,512,0.009569777382744683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,768,0.04418311185306973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,256,0.007336888876226213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,768,0.07029689020580716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,512,0.032264888286590576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,512,0.05311200022697449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,256,0.023185777995321486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,128,0.006093333164850871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,256,0.040101332796944514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,128,0.01926399932967292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,64,0.00545155546731419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,32,0.005626666463083691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,64,0.0180711117055681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,128,0.033992889854643084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,32,0.017669333351982962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,16384,0.11467644903394912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,16384,0.8074720170762805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,12288,0.08817244238323635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,16384,1.0031351513332791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,12288,0.6123298009236654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,65536,0.44726933373345273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,10240,0.07426133420732287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,12288,0.7601635720994738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,8192,0.05976000097062853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,10240,0.5210071139865452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,10240,0.6395733091566298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,8192,0.41310845481024844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,7168,0.05349066522386339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,8192,0.517573356628418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,6144,0.04610400067435371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,7168,0.36385244793362087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,65536,3.172168943617079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,5120,0.04054488738377889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,7168,0.45620001686943906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,6144,0.314721769756741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,4096,0.03299644589424133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,6144,0.39693334367540145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,5120,0.2652026547325982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,3584,0.02920888861020406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,5120,0.3340151045057509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,4096,0.21602488888634575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,3072,0.025776889589097764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,3584,0.1899884409374661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,65536,3.9443316989474826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,4096,0.27450309859381783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,3072,0.16471378008524576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,3584,0.24222932921515572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,2560,0.022626666559113398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,2048,0.01846844454606374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,2560,0.13862044281429714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,3072,0.21090222729576957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,1536,0.014843554960356818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,2048,0.11241866482628717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,2560,0.1805244419309828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,1024,0.011335111326641507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,1536,0.08031200038062201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,2048,0.1468764411078559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,1536,0.11134755611419678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,768,0.00962222201956643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,1024,0.053526222705841064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,512,0.007841777470376756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,1024,0.08305422465006511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,768,0.04367555512322319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,256,0.006245333287451003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,512,0.03272533416748047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,768,0.06858044200473361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,512,0.05376177695062426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,128,0.0053075556125905775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,256,0.023188443647490606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,256,0.039157334301206805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,128,0.019401777121755812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,64,0.0046826667255825466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,32,0.004899555610285865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,128,0.03340444299909804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,64,0.018226666582955253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,32,0.017662222186724346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,16384,0.07912177509731717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,16384,0.7662320137023926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,65536,0.3072062333424886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,12288,0.06116000148985121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,12288,0.5801946851942275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,128,0.034683555364608765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,10240,0.052278220653533936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,65536,3.020064036051432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,256,0.04101244608561198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,8192,0.0467582212554084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,10240,0.4882693290710449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,7168,0.038895110289255776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,8192,0.39371199078030056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,7168,0.3466826544867621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,6144,0.03397599856058756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,6144,0.30094043413798016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,512,0.05667199691136678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,5120,0.028714666763941448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,5120,0.25311022334628636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,16384,0.9936808480156792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,4096,0.024217777782016333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,4096,0.2061786651611328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,12288,0.7540888786315918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,3584,0.02075466679202186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,3584,0.18181333276960585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,8192,0.511947578854031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,3072,0.018263111511866253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,3072,0.15830043951670328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,65536,3.8879661560058594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,10240,0.6346053547329372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,7168,0.45239464441935223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,2560,0.016000888413853116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,2048,0.013565333353148567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,2560,0.13240977128346762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,2048,0.10743021965026855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,1536,0.010909332997269101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,1536,0.07694844404856364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,6144,0.3910008801354303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,1024,0.008449777960777283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,1024,0.05157866742875841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,5120,0.3299955526987712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,768,0.007294221884674496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,768,0.041401777002546526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,4096,0.27024798923068577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,512,0.006248000181383557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,512,0.03067644437154134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,3584,0.23901687728034127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,256,0.00517511119445165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,256,0.02167822255028619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,2560,0.17719644970364043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,3072,0.20815200275845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,128,0.004511111312442356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,128,0.018212444252438016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,64,0.0041422223051389056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,64,0.01682222220632765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,32,0.004268444246715969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,32,0.01645511057641771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,1536,0.11027822229597305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,2048,0.14426399601830378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,1024,0.08081155353122287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,16384,0.06446844339370728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,65536,0.22306667433844674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,16384,0.757114675309923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,12288,0.050494223833084106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,768,0.06831288999981351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,12288,0.5735084215799967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,10240,0.04670133193333944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,65536,2.96870125664605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,10240,0.48127465777926975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,8192,0.032552000549104475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,8192,0.3894142309824626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,256,0.03829866647720337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,7168,0.03103289008140564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,512,0.05145600106981066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,128,0.032468444771236844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,7168,0.3434435526529948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,6144,0.028026666906144884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,6144,0.29738577206929523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,5120,0.024843555357721116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,5120,0.25024178293016225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,16384,0.9775706397162544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,4096,0.01941422290272183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,4096,0.20255022578769258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,12288,0.738458686404758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,3584,0.01718844473361969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,3584,0.17889422840542266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,10240,0.6232222451104058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,3072,0.01569333341386583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,3072,0.1556604438357883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,65536,3.816486358642578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,8192,0.5009751319885254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,2560,0.013277333643701343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,2560,0.13150310516357422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,2048,0.011256888508796692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,2048,0.10556977987289429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,7168,0.4432231055365668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,1536,0.00927111092540953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,1536,0.075436446401808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,6144,0.38582221666971844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,1024,0.0073315559162033936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,1024,0.0507262216673957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,5120,0.32635556326972115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,768,0.006188444379303191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,768,0.040956444210476346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,4096,0.2663662168714735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,512,0.0052986666560173035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,512,0.029319110843870375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,3584,0.2349031236436632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,256,0.0041511112617121804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,256,0.021173333128293354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,3072,0.20494755109151205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,128,0.003752888904677497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,128,0.017708443933063082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,2560,0.1740551127327813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,64,0.0034560000316964257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,64,0.01626133256488376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,32,0.0035377778112888336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,32,0.015779554843902588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,65536,0.20526222387949625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,65536,2.955297682020399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,2048,0.1422462198469374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,16384,0.06079644627041287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,16384,0.755943086412218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,1536,0.10799733135435317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,12288,0.04006577862633599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,768,0.06685066885418363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,12288,0.5709617932637533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,10240,0.034479998879962497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,1024,0.07979910903506808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,10240,0.47940535015530056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,8192,0.02776355544726054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,8192,0.38751999537150067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,256,0.03687644335958693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,7168,0.025123554799291823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,7168,0.34144266446431476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,512,0.051791111628214516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,6144,0.022476444641749065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,128,0.03127200073666043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,6144,0.2956791188981798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,5120,0.018283555905024212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,5120,0.24955556127760145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,16384,0.9733235041300455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,4096,0.010701333483060202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,4096,0.20279555850558809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,65536,3.8043645222981772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,3584,0.009883555273214975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,3584,0.17875999874538848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,12288,0.7377920150756836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,3072,0.00905955582857132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,3072,0.15523466798994276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,10240,0.6189813084072536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,2560,0.007801777786678738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,2560,0.13058311409420437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,8192,0.5011244350009495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,2048,0.006794666664467917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,2048,0.10502844386630589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,7168,0.4407884544796414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,1536,0.005948444621430502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,6144,0.38166220982869464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,1536,0.07478044430414836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,1024,0.004968889057636261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,1024,0.04999022351370918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,5120,0.324672884411282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,768,0.004654222064548069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,768,0.040828443235821195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,4096,0.2639422151777479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,512,0.004008888784382078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,512,0.02902666727701823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,3584,0.23260622554355195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,256,0.003600888782077365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,256,0.020982222424613103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,3072,0.20304621590508354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,128,0.003338666632771492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,128,0.01741422216097514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,2560,0.17318311002519396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,64,0.0032044444233179092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,64,0.016114667057991028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,32,0.0033937777496046494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,32,0.015607110328144498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,65536,0.19507643911573622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,2048,0.14084445105658636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,16384,0.0536480016178555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,16384,0.7548186514112684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,1536,0.1069564421971639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,12288,0.042803555727005005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,10240,0.036529776122834944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,1024,0.07800533374150594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,12288,0.571441756354438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,10240,0.4795840051439073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,8192,0.0306328899330563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,7168,0.027898665931489732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,8192,0.3879351086086697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,6144,0.02163377735349867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,7168,0.3420853349897597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,768,0.0653688907623291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,65536,2.9594294230143228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,5120,0.016678222351604037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,4096,0.0090515555606948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,3584,0.008282666405042013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,6144,0.29612000783284503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,5120,0.24960621198018393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,3072,0.008995555341243744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,4096,0.20249688625335693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,2560,0.007872000336647034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,3584,0.17908622158898246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,2048,0.0068711112770769335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,3072,0.15483910507626003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,1536,0.005818666683302985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,2560,0.13046133518218994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,1024,0.004828444371620814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,2048,0.10510666502846612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,1536,0.0747671127319336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,768,0.004361777669853634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,1024,0.05023466547330221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,512,0.00398488880859481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,768,0.04022488991419474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,256,0.0034711110509104202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,512,0.029338667790095013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,128,0.003294222263826264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,256,0.020979555116759405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,64,0.003152888889114062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,128,0.017287110288937885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,32,0.0031866667171319327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,64,0.016043555405404832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,32,0.015431111057599386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,16384,0.051982220676210195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,65536,0.19176177183787027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,12288,0.04118400149875217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,10240,0.03529333406024509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,12288,0.5717351171705458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,16384,0.7556871308220757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,8192,0.02918755677011278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,10240,0.47992976506551105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,7168,0.026510222090615168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,6144,0.023800000548362732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,8192,0.38837867312961155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,5120,0.015666666958067153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,7168,0.342320892545912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,4096,0.00812533332241906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,6144,0.29551733864678276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,3584,0.007648000286685095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,3072,0.007062222394678328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,5120,0.24962221251593697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,4096,0.2028568850623237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,2560,0.00638488887084855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,3584,0.17881955040825737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,2048,0.005613333235184352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,3072,0.15476622846391466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,2560,0.13065244091881648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,1536,0.005056888692908817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,2048,0.1049777799182468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,1024,0.004338666796684265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,1536,0.07500977648629083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,768,0.004024888906213972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,1024,0.05016533202595181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,512,0.0036657779581016968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,768,0.040255109469095864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,256,0.0033644443998734155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,512,0.02921688887808058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,128,0.003107555624511507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,256,0.020997333857748244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,64,0.0029075555503368378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,128,0.0174319992462794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,32,0.003088888815707631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,64,0.01604977746804555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,32,0.015418666932317944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,65536,2.9572647942437063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,256,0.03641600079006619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,512,0.04935377836227417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,12288,2.7722471025254993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,16384,3.53202395968967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,10240,3.3351501888699002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,10240,2.3172844780815973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,12288,4.031152937147352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,16384,5.285542382134332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,8192,1.8400675455729167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,128,0.030918220678965252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,8192,2.6987342834472656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,12288,2.661134295993381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,7168,1.6231760448879664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,16384,3.767319997151693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,6144,1.3839155832926433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,6144,2.032311121622721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,5120,1.144927978515625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,7168,2.3388356102837458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,10240,2.3186143239339194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,5120,1.6864746941460504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,4096,0.918518172370063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,4096,1.372947586907281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,3584,0.7987573411729602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,8192,1.8621741400824652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,3584,1.2081226772732205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,3072,0.6774009068806967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,7168,1.5623244179619684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,3072,1.0264800389607747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,2560,0.5578586790296767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,2560,0.7967333263821073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,2048,0.6669448746575249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,2048,0.4388337665134006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,6144,1.3780541949801977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,5120,1.1099626753065321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,1536,0.3343671162923177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,1536,0.49246576097276473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,4096,0.9358355204264323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,1024,0.22643733024597168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,1024,0.348399109310574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,768,0.18684532907274035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,768,0.2781093385484484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,3584,0.7833262019687228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,512,0.20882666110992432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,512,0.15346044964260525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,256,0.14050755235883924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,256,0.13418043984307185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,3072,0.6624302334255642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,128,0.11716711521148682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,128,0.12498844994439019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,2560,0.5739804373847114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,64,0.11659910943773057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,64,0.1321457756890191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,32,0.11289600531260173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,32,0.11949511369069417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,1536,0.3756995466020372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,256,0.14327111509111193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,2048,0.46906222237481016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,512,0.17454311582777235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,768,0.23131023512946236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,1024,0.27933422724405926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,16384,1.1530141830444336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,128,0.14631466070810953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,16384,1.2662329143948026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,16384,1.2149555418226454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,12288,0.8957537545098199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,12288,0.9122240278455945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,12288,0.9857244491577148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,10240,0.7582151095072428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,8192,0.5922115643819174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,10240,0.7418168915642632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,10240,0.7675520049201118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,7168,0.5210311147901747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,8192,0.5982213550143772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,8192,0.6156711048550075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,65536,5.013340420193143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,7168,0.5297270880805122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,7168,0.5459493531121148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,6144,0.4542444547017415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,6144,0.4699661996629503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,5120,0.3740533457862006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,6144,0.45478132035997176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,5120,0.3780248959859212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,5120,0.40048800574408633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,4096,0.30551645490858287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,3584,0.2600071165296766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,4096,0.3281039926740858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,65536,4.988668653700087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,4096,0.306151098675198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,3584,0.27365599738227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,3584,0.28672533565097386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,3072,0.22048710452185738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,3072,0.25076177385118276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,2560,0.18037777476840547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,2560,0.21085688802931044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,2560,0.20254933834075928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,2048,0.1448462274339464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,3072,0.23765778541564941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,1536,0.11011466715070937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,2048,0.18082577652401396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,2048,0.16530666086408827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,1024,0.08084799846013387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,1536,0.1453288926018609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,768,0.07441777653164335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,768,0.06875022252400716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,1024,0.10928889115651448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,512,0.05667110946443346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,1024,0.09157333109113906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,256,0.037351999017927386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,512,0.05647822221120199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,512,0.07503022087944879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,128,0.032807111740112305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,256,0.04745777779155307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,1536,0.12666222784254286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,768,0.09434488746854995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,256,0.06184444162580702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,64,0.030119998587502375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,128,0.047135998805363975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,128,0.059954669740464955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,32,0.030926220946841772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,64,0.044785777727762856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,32,0.04385688900947571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,16384,0.9543351067437066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,16384,1.0261138280232747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,12288,0.731237358517117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,16384,0.9335199991861979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,12288,0.7382284270392524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,65536,5.146979437934028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,12288,0.774072011311849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,10240,0.6198746893141005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,10240,0.6534586482577854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,8192,0.4935750961303711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,10240,0.575849797990587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,8192,0.4534782303704156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,8192,0.5249839888678657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,7168,0.4079013400607639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,7168,0.43319198820326066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,65536,3.9281340705023875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,6144,0.3758702278137207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,7168,0.46503554450141055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,6144,0.3391297658284505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,6144,0.40515822834438747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,5120,0.3106035656399197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,5120,0.338551123936971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,5120,0.28769421577453613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,4096,0.2368417845831977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,4096,0.27915999624464244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,4096,0.2605280081431071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,65536,4.246148427327474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,3584,0.21657778157128227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,3584,0.20500622855292427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,3072,0.18123555183410645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,2560,0.1486391093995836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,2560,0.18017333083682588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,3584,0.2441013389163547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,2048,0.12029600143432617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,3072,0.21325955126020643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,3072,0.17941421932644316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,1536,0.0929573310746087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,2560,0.1559822161992391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,2048,0.15141599708133274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,1536,0.12521778212653265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,2048,0.1249048842324151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,1024,0.06893333461549547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,1024,0.07102044423421223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,768,0.056700441572401256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,768,0.058989332781897656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,1024,0.09863466686672634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,512,0.043657776382234365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,1536,0.09801689121458267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,768,0.08261511060926649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,256,0.0299253331290351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,128,0.025589333640204534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,512,0.04826577835612827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,256,0.04043022129270766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,512,0.06619377930959065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,256,0.05605777766969469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,64,0.024271999796231587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,128,0.03642844491534763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,64,0.03874844312667847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,32,0.025437333517604407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,128,0.05307911170853508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,32,0.03751022285885281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,16384,0.8477884398566352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,12288,0.5934506522284614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,16384,0.9335412979125977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,12288,0.6586755646599664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,65536,3.8676906161838107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,16384,0.7410657670762805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,12288,0.7029973136054145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,10240,0.5551377932230631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,10240,0.47359646691216367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,10240,0.5958764288160536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,8192,0.4390675491756863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,8192,0.4876409106784397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,8192,0.3823920090993245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,7168,0.38544355498419863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,7168,0.42456534173753524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,6144,0.3356364568074544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,7168,0.3394808769226074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,6144,0.2917626698811849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,6144,0.37091732025146484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,5120,0.27885601255628795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,5120,0.2458826700846354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,4096,0.22450666957431367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,5120,0.3128071096208361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,4096,0.19797066847483316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,3584,0.1931075519985623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,4096,0.2552986674838596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,3584,0.17677422364552817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,3072,0.16144266393449572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,3584,0.22462044821845162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,3072,0.19290222062004936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,3072,0.15362666712866888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,2560,0.1328622235192193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,2048,0.10671644740634495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,2560,0.16730222437116835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,65536,3.434071011013455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,65536,3.7489670647515188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,2048,0.14047111405266657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,2560,0.1309386624230279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,1536,0.0849999984105428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,1536,0.0842942198117574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,1024,0.06604355573654175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,1024,0.06281510988871257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,768,0.04972177743911743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,768,0.05390310949749417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,2048,0.10819377501805623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,1024,0.088082664542728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,1536,0.1130488846037123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,512,0.03867111272282071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,768,0.07600888941023085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,512,0.04422044422891405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,256,0.027644443843099806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,512,0.06242844131257799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,256,0.036432888772752546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,128,0.022052443689770166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,256,0.05187199844254387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,128,0.03285600079430474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,64,0.019620445039537217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,32,0.020226667324701946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,128,0.04890133274926079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,64,0.033858666817347206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,32,0.03340889016787211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,16384,0.7617662217881945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,16384,0.8569119771321615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,16384,0.6069022284613715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,12288,0.44672086503770614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,12288,0.5838328997294108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,12288,0.6488018035888672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,10240,0.48970312542385525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,10240,0.37299378712972003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,8192,0.3060426712036133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,10240,0.5344008869594997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,8192,0.38710933261447483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,8192,0.4348168902926975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,7168,0.3416853215959337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,65536,3.3358489142523875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,7168,0.26395021544562447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,7168,0.38627200656467015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,6144,0.23116532961527506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,6144,0.2973537709977892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,5120,0.24590667088826498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,5120,0.19437244203355578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,6144,0.3314017719692654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,4096,0.19913600550757515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,5120,0.28345778253343373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,65536,3.034412384033203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,3584,0.17114844587114122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,4096,0.15668711397382948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,3584,0.2038053274154663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,4096,0.23403554492526582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,65536,3.313251495361328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,3072,0.1422817839516534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,3072,0.1751964489618937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,3584,0.1414382192823622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,2560,0.11742844846513535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,2048,0.09536977608998616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,2560,0.15447021855248347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,1536,0.06684088706970215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,2048,0.1284631093343099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,2560,0.10474844773610432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,3072,0.12139288584391277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,1536,0.07495466868082683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,1024,0.04895200000868904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,2048,0.08536444769965278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,1024,0.056277334690093994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,1536,0.10548533333672418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,768,0.040778666734695435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,512,0.03105777833196852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,768,0.048284444544050426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,256,0.021574222379260596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,1024,0.08126488659116957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,512,0.03978577918476529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,768,0.07025955782996283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,256,0.032441778315438166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,256,0.04897333184878031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,512,0.059009777175055615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,128,0.016733333468437195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,128,0.029286222325430974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,64,0.014703111516104804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,32,0.015969778100649517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,128,0.04459111226929558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,64,0.029570665624406602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,32,0.028897778855429754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,16384,0.7061111132303873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,12288,0.4180702103508844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,16384,0.5514026747809516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,12288,0.5419164233737522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,65536,2.608280817667643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,10240,0.45568354924519855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,10240,0.3562515576680501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,8192,0.2742328908708361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,8192,0.3628071149190267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,65536,2.8238249884711375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,7168,0.31694044007195366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,7168,0.24960978825887045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,6144,0.21309867170121935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,6144,0.27751289473639595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,65536,2.2013022104899087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,5120,0.17933155430687797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,5120,0.2312542332543267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,16384,0.8464168972439237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,4096,0.14094221591949463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,12288,0.6447999742296007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,4096,0.18482222821977404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,3584,0.1243164406882392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,3584,0.1593422227435642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,10240,0.5446231100294325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,3072,0.108115553855896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,3072,0.1339555581410726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,8192,0.440465768178304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,65536,3.382569842868381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,2560,0.11027733484903972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,2560,0.09003288878334893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,2048,0.07369777891370985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,2048,0.08940622541639541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,7168,0.38995199733310276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,1536,0.056464003192053906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,1536,0.07087644603517321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,6144,0.3382008870442708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,1024,0.040503111150529646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,1024,0.05333155393600464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,5120,0.28756711218092174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,768,0.03261333372857835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,768,0.046134223540623985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,4096,0.2348408963945177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,512,0.024774221910370722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,3584,0.20739377869500053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,512,0.03787111242612203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,256,0.01757688820362091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,256,0.030696001317765977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,2560,0.15540266036987305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,128,0.014567111929257711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,128,0.027151111099455092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,1536,0.10586755805545384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,64,0.013567999833159976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,64,0.027811557054519653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,32,0.014993778533405729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,32,0.026349332597520616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,3072,0.17775911755032012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,2048,0.12886844740973577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,1024,0.08386488755544026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,768,0.07129955291748047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,65536,2.595081753200955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,16384,0.6525013181898329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,16384,0.44489865832858616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,16384,0.751796457502577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,65536,1.9531741672092016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,65536,2.8994994693332248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,256,0.04900977677769131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,12288,0.33465509944491917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,12288,0.5067653126186794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,12288,0.5752959781222874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,10240,0.42310667037963867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,512,0.05984533495373196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,10240,0.28052179018656415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,10240,0.48053868611653644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,8192,0.33849867184956867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,128,0.04435733291837904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,8192,0.23011467191908094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,8192,0.3948800033993191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,7168,0.29483376608954537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,7168,0.34654490152994794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,5120,0.213027556737264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,6144,0.25784799787733287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,6144,0.3037422233157688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,6144,0.17275555928548178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,5120,0.25500443246629506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,7168,0.20165244738260904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,4096,0.17237422201368544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,3584,0.1476071145799425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,4096,0.2074862188763089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,3072,0.12355732917785645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,3584,0.18304977152082655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,5120,0.14816799428727892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,2560,0.07978310849931505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,4096,0.11917244063483344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,3584,0.10580178101857503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,3072,0.1574337747361925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,2560,0.10314755307303534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,2048,0.06647021902932061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,3072,0.09269155396355523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,1536,0.05189244283570183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,2048,0.0835804475678338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,1536,0.06634489032957289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,2560,0.13536977767944336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,1024,0.038937777280807495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,2048,0.11572710673014323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,1536,0.09610399934980605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,1024,0.050309333536359996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,768,0.03191377719243368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,1024,0.07384177711274889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,768,0.04345600141419304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,512,0.024578667349285547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,512,0.03551466597451104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,768,0.06455555227067736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,256,0.017786666750907898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,512,0.05467555589146084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,256,0.028608888387680054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,128,0.013886222408877479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,32,0.01275466630856196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,256,0.04424266682730781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,128,0.02476177778508928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,64,0.011931555966536203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,64,0.024554666545655992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,128,0.04015644391377767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,32,0.024166221419970196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,16384,0.6043981975979275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,16384,0.3712933328416612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,12288,0.2829066647423638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,12288,0.4625386661953396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,10240,0.3913919925689697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,10240,0.23745510313245985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,8192,0.3109644518958198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,8192,0.19206844435797799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,7168,0.169121781984965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,65536,2.3918889363606772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,7168,0.2731404569413927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,6144,0.14620710743798151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,6144,0.23811732398139107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,65536,1.7190728717380102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,5120,0.12255999777052139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,5120,0.19783466392093232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,16384,0.7520844671461316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,4096,0.10090666347079807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,4096,0.16010133425394693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,12288,0.5757306416829427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,3584,0.08911199702156915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,3584,0.13694932725694445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,10240,0.4820906851026747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,3072,0.07759022050433688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,3072,0.11399111482832168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,8192,0.3940444522433811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,2560,0.0665466652976142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,2560,0.0946764416164822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,65536,2.9709146287706165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,2048,0.057816889550950795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,7168,0.35083733664618594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,2048,0.07704355319341023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,1536,0.044030222627851695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,1536,0.06216533316506279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,6144,0.3027884430355496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,1024,0.03228088882234361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,1024,0.04742844568358528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,5120,0.25797422726949054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,768,0.025987555583318073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,768,0.04074755642149184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,3584,0.18617688284979927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,512,0.020023999942673575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,512,0.03318488929006789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,4096,0.2085120015674167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,256,0.014171555638313293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,256,0.026608000199000042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,2560,0.13718310991923013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,128,0.01069155583779017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,128,0.02331377731429206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,1536,0.0958942241138882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,64,0.009152000149091085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,64,0.02249777813752492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,32,0.009431111315886179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,32,0.022015111313925848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,3072,0.1588968833287557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,2048,0.11578933397928874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,768,0.06533688969082303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,16384,0.29707733790079754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,16384,0.5536417961120605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,65536,2.195343017578125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,1024,0.07514933082792494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,16384,0.6634791162278917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,12288,0.42646132575141055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,65536,1.2524355782402885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,12288,0.22374134593539766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,12288,0.5052675671047634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,65536,2.5417679680718317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,10240,0.3582319948408339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,10240,0.18719022803836396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,10240,0.42560532357957626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,8192,0.15359555350409612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,8192,0.28623912069532603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,8192,0.3491528828938802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,7168,0.2514044443766276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,7168,0.30838221973843044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,6144,0.21798045105404326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,7168,0.1334808932410346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,6144,0.2686844401889377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,5120,0.18180621994866264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,5120,0.09888711240556504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,5120,0.22653243276807997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,6144,0.11604444185892741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,4096,0.14661422040727404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,3584,0.07039199935065375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,4096,0.1863795518875122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,3584,0.1265546613269382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,4096,0.08150577545166016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,3584,0.16410666041904026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,3072,0.06142933501137627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,3072,0.10479111141628689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,2560,0.05394666724734836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,3072,0.14155644840664333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,2560,0.08590755860010783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,2048,0.044575111733542554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,2048,0.07060799996058147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,2560,0.11972622076670329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,1536,0.03499822152985467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,2048,0.10124088658226861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,1536,0.05748089154561361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,1024,0.02516799999607934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,1536,0.085917333761851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,1024,0.04189777705404493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,768,0.020678222179412842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,512,0.016185777054892648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,1024,0.06432622008853488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,768,0.03801066676775614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,768,0.05856621927685208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,512,0.02927910950448778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,256,0.011798222031858234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,512,0.04741155438952976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,128,0.008822222550710043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,256,0.023440000083711412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,256,0.03824177715513442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,128,0.02067466742462582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,64,0.00757422215408749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,128,0.03550844391187032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,32,0.008056888977686564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,32,0.020655110478401184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,64,0.02051822178893619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,512,0.054281777805752225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,256,0.04387377699216207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,16384,0.2965022193060981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,16384,0.5513600243462456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,128,0.0414266652531094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,12288,0.4211475584242079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,12288,0.5081306563483344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,16384,0.6638533274332682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,12288,0.2249208821190728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,10240,0.35544710689120823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,8192,0.28301334381103516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,10240,0.18823911084069145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,10240,0.4297777811686198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,8192,0.15054221947987875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,7168,0.2488408883412679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,8192,0.34954577022128636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,6144,0.2166888846291436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,7168,0.3105564382341173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,7168,0.13343911700778538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,65536,2.1623378329806857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,6144,0.26875556839836967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,6144,0.11475377612643772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,5120,0.18119911352793375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,4096,0.07985955476760864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,4096,0.14510933558146158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,5120,0.22779644860161674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,3584,0.07130577829149035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,3584,0.12523555755615234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,4096,0.18513866265614828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,3072,0.06152088774575127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,5120,0.09827911191516453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,3584,0.16587911711798772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,3072,0.10510666502846612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,2560,0.052377776967154614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,2560,0.08652889066272312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,3072,0.14051911565992567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,2048,0.04435733291837904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,2560,0.12261333730485703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,2048,0.06957333617740206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,1536,0.03443377878930833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,1536,0.05715733104281955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,1024,0.025495110286606684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,1536,0.08521955543094212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,2048,0.10197955369949341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,1024,0.041747556792365186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,768,0.03772088885307312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,768,0.020614221692085266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,1024,0.06567555665969849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,512,0.016193777322769165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,768,0.05880800220701429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,256,0.011325333681371478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,512,0.049341334236992725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,512,0.029623111089070637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,256,0.023131555981106226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,128,0.008525333470768398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,256,0.03978844483693441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,64,0.00740799970097012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,128,0.020448888341585796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,128,0.03717866539955139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,32,0.008069333102968004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,65536,2.583896001180013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,64,0.020463110672103036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,32,0.020310221446885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,12288,0.17053688897026908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,16384,0.5084106657240126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,16384,0.22773067156473795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,65536,1.2340106964111328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,10240,0.32778043217129177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,12288,0.3889519903394911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,10240,0.14531555440690783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,8192,0.2622213363647461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,8192,0.11586577362484402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,7168,0.10225244363149007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,7168,0.23057778676350912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,65536,1.9919698503282335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,6144,0.08878311183717515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,6144,0.20101955201890734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,65536,1.049497816297743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,5120,0.07597600089179145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,5120,0.16772622532314727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,16384,0.6167110866970485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,4096,0.06203111012776693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,4096,0.13554755846659342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,8192,0.3252986537085639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,12288,0.4713137944539388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,3584,0.055121776130464345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,10240,0.39506133397420246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,3584,0.11642666657765706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,3072,0.048226667775048145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,3072,0.09727199872334798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,2560,0.040960888067881264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,2560,0.08020799689822726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,7168,0.28515201144748265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,2048,0.034161776304244995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,2048,0.06461777951982287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,6144,0.2494035561879476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,1536,0.027246221899986267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,65536,2.3580968644883895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,1536,0.05219733383920458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,1024,0.020053333706325956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,1024,0.03932977716128031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,5120,0.2121431165271335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,768,0.01663377715481652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,768,0.03409333361519708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,3584,0.15265956189897326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,512,0.013136000268989138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,512,0.026833777626355488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,2560,0.11292621824476455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,256,0.009054222040706212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,256,0.02161600026819441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,4096,0.17347821924421522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,128,0.007296889192528195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,128,0.019388443893856473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,768,0.05322133170233833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,64,0.006367110957702001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,3072,0.13114133146074083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,64,0.019340443942281935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,32,0.0069520001610120135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,32,0.018975110517607797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,1536,0.0796862244606018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,2048,0.09538844558927749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,1024,0.06094488832685682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,65536,1.957562764485677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,16384,0.18712088796827528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,16384,0.5006648699442545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,12288,0.14204711384243435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,12288,0.3835795455508762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,65536,0.8383466402689616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,10240,0.12025066216786702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,10240,0.3271733389960395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,8192,0.09672888782289292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,8192,0.257168001598782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,128,0.033979554971059166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,7168,0.0868871079550849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,256,0.03625955515437656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,7168,0.22814665900336373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,6144,0.07551911142137316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,6144,0.19788622856140137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,512,0.04551377892494202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,5120,0.0632737808757358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,5120,0.16507466634114584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,16384,0.5992950863308376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,4096,0.05199288990762499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,4096,0.13392178217569986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,12288,0.4558177524142795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,3584,0.045891556474897593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,65536,2.2890667385525174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,10240,0.385151121351454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,3584,0.11549866199493408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,3072,0.04056444433000352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,3072,0.09577688905927871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,2560,0.034501334031422935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,2560,0.0793608890639411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,8192,0.3152257866329617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,2048,0.02879022227393256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,2048,0.06447111235724555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,7168,0.2795271078745524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,1536,0.023285332653257582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,1536,0.051903999514049955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,6144,0.24255111482408312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,1024,0.017439110411538016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,1024,0.039114667309655085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,5120,0.20728000005086264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,768,0.014304889572991265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,768,0.03336266676584879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,2560,0.11076888773176406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,512,0.011370666325092316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,4096,0.16912266943189833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,512,0.026271111435360376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,256,0.008378666308191087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,256,0.021386666430367365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,3584,0.1495493385526869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,128,0.007111999723646376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,128,0.01942933268017239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,3072,0.12872000535329184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,64,0.006121777825885349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,64,0.01928711103068458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,32,0.006423111177153057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,32,0.01887911061445872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,2048,0.09365066554811265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,768,0.05199999941719902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,65536,1.8122452629937067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,65536,0.5836728943718804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,1536,0.07805244127909343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,65536,2.1881910959879556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,16384,0.46364180246988934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,16384,0.14889421727922228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,12288,0.3549511167738173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,16384,0.5733226670159234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,12288,0.11302310890621609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,12288,0.43622401025560165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,10240,0.09595644474029541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,10240,0.29957956737942165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,10240,0.3720888826582167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,8192,0.23905687861972383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,8192,0.0772577789094713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,7168,0.06929422087139554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,8192,0.30116356743706596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,6144,0.06029866801367866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,7168,0.21103021833631727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,6144,0.182824002371894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,7168,0.26640711890326607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,5120,0.050833778248892895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,6144,0.23434933026631674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,4096,0.041803555356131665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,5120,0.1529902219772339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,5120,0.1964071061876085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,3584,0.036990222003724836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,4096,0.12348711490631104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,4096,0.16107911533779568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,3072,0.03288622366057502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,3584,0.10640711254543728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,3584,0.14053689108954534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,3072,0.08833689159817165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,2560,0.027841776609420776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,3072,0.12307466400994195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,2048,0.02344177828894721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,2560,0.07333333624733819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,2048,0.05940710835986667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,2560,0.10518488619062637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,1536,0.018851555056042142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,2048,0.08848977751202053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,1024,0.01425155500570933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,1536,0.047241777181625366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,1024,0.03426222337616815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,1536,0.07430399788750543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,768,0.01199288914601008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,1024,0.06022666560278999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,768,0.029138667715920344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,1024,0.05723910861545139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,512,0.009549332989586724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,768,0.05014666583802965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,512,0.02420266634888119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,256,0.007127111156781514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,512,0.04096533192528619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,256,0.01977777812216017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,128,0.006120000034570694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,256,0.034640000926123724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,128,0.017747556169827778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,64,0.0053377776510185665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,32,0.005681777579916849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,128,0.032778667079077825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,64,0.01780177818404304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,32,0.017449777987268236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,16384,0.45656797620985246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,16384,0.1151884396870931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,16384,0.5494364632500542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,12288,0.08812710973951553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,12288,0.3489546775817871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,10240,0.0732497771581014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,12288,0.4200337727864583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,10240,0.29380978478325737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,8192,0.05986399783028496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,65536,1.7803192138671875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,10240,0.35377687878078884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,65536,0.445250670115153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,8192,0.23547734154595268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,7168,0.053263111246956714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,6144,0.04646755589379204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,7168,0.20987466971079508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,8192,0.2898168828752306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,7168,0.25719467798868817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,5120,0.039431111680136785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,6144,0.18050932884216309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,4096,0.03289866778585646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,6144,0.22435998916625977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,5120,0.15135822031233045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,65536,2.0998995039198136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,5120,0.1895057757695516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,3584,0.028993778758578833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,4096,0.12238311767578125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,128,0.033392889632119074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,3072,0.025553777813911438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,4096,0.1547022263209025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,3584,0.10546400149663289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,2560,0.022076444493399724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,3072,0.08804711368348862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,3584,0.1350204414791531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,2048,0.01846311158604092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,2560,0.07207644648022123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,3072,0.1181902223163181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,2560,0.1019973357518514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,1536,0.015306666493415833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,2048,0.058556444115108915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,2048,0.08567288849088882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,1024,0.011651555697123209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,1536,0.04756444361474779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,1024,0.035092443227767944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,768,0.00962933318482505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,1536,0.07206222083833483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,1024,0.05601244502597385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,512,0.007811555431948767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,768,0.03094844354523553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,768,0.049514667855368726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,256,0.00609777785009808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,512,0.024726221958796184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,512,0.040017777019076876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,128,0.00518488883972168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,256,0.020152888364262052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,256,0.03375555409325494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,128,0.018068444397714403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,64,0.004659555438492033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,128,0.03202933404180739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,32,0.004876444323195351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,64,0.017813333206706576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,32,0.01759822169939677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,16384,0.41727201143900555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,256,0.03709333472781711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,16384,0.0776453349325392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,12288,0.05843377775616116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,12288,0.32213687896728516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,10240,0.052389333645502724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,10240,0.2718568907843696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,8192,0.04138400157292684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,65536,0.30603822072347003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,8192,0.21947733561197916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,7168,0.0365866654449039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,65536,1.6109004550509984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,7168,0.1938720014360216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,6144,0.03184177809291416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,6144,0.16780266496870253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,512,0.04453511039415995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,5120,0.027450667487250432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,5120,0.14002933767106798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,16384,0.5321795675489638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,12288,0.4079333411322699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,4096,0.02277244461907281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,4096,0.11284000343746609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,3584,0.020185778538386028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,3584,0.09757777717378403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,10240,0.3453048865000407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,8192,0.2805315653483073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,3072,0.018182223041852314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,3072,0.08081689145829943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,2560,0.015634665886561077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,65536,2.0377928415934243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,2560,0.06673066483603583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,2048,0.013392888837390475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,2048,0.054924445019827954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,7168,0.24843200047810873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,1536,0.010858666565683154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,1536,0.0439280006620619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,6144,0.2174382209777832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,1024,0.008366222182909647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,1024,0.033450666401121355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,5120,0.18348977300855848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,768,0.007285333342022366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,768,0.028353777196672227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,4096,0.14978310796949598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,512,0.006182222316662471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,3584,0.13134755028618708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,512,0.022488888767030504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,256,0.005024000174469418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,256,0.01854222185081906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,3072,0.11376799477471246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,128,0.004446222136418025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,128,0.016737777325842116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,2560,0.09817777739630805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,64,0.004020444634887907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,64,0.016262221667501662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,32,0.004274666723277834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,32,0.016044444508022733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,2048,0.08324266804589166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,1536,0.07030400302675036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,1024,0.05470310979419284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,16384,0.08244444264305963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,65536,0.23681865798102486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,16384,0.41095466083950466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,12288,0.06051555607053968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,65536,1.5934036042955186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,12288,0.31428978178236217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,10240,0.054139554500579834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,10240,0.26510845290289986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,768,0.04719466633266873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,8192,0.033379554748535156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,8192,0.21239999930063883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,256,0.033067554235458374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,128,0.031002667215135362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,7168,0.03614844547377692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,512,0.0388426681359609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,7168,0.18877600298987496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,6144,0.03242666522661845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,6144,0.16406667232513428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,5120,0.02718577782313029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,5120,0.13814755280812582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,16384,0.5315386454264323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,4096,0.019371555911170114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,4096,0.11177689499325222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,12288,0.4054639869266086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,3584,0.017683555682500202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,3584,0.09592888752619426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,65536,2.029491636488173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,3072,0.01609777741962009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,10240,0.34404622183905703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,3072,0.08027555545171101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,2560,0.013583111266295115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,2560,0.06610666380988227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,8192,0.2788871129353841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,2048,0.011490667031870948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,2048,0.052932444545957774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,7168,0.24672444661458334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,1536,0.009474666582213508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,1536,0.04296000136269463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,6144,0.21550133493211535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,1024,0.007296889192528195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,1024,0.032066666417651706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,5120,0.18285511599646675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,768,0.0063822223908371395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,768,0.02739199995994568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,3584,0.13108266724480525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,512,0.005355555564165115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,512,0.022146667043368023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,2560,0.09768178065617879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,256,0.004548444516128964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,256,0.018343999981880188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,4096,0.14876355065239802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,128,0.004032888760169347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,128,0.016372443901167977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,3072,0.11352088716295029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,64,0.003719999972316954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,64,0.016121778223249648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,2048,0.08301333586374919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,32,0.003933333274390963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,32,0.01574222246805827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,65536,0.16285155879126653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,65536,1.5585653516981337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,16384,0.05123733480771383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,16384,0.40538043446011013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,1536,0.07022044393751356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,12288,0.04842311143875122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,768,0.04743644595146179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,12288,0.30866577890184194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,10240,0.04253866606288486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,10240,0.26036445299784344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,1024,0.05404622356096903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,8192,0.030448890394634668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,8192,0.21221689383188883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,128,0.030449777841567993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,7168,0.02960088849067688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,256,0.03230666783120897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,7168,0.1866382228003608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,6144,0.025951999757024977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,6144,0.16161333190070257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,512,0.038196444511413574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,5120,0.02330933345688714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,5120,0.13614133993784586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,16384,0.5174009005228678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,4096,0.0173999998304579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,4096,0.1098160015212165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,65536,1.985073725382487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,3584,0.0162764439980189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,3584,0.09495644436942206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,12288,0.39733065499199766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,3072,0.01475911173555586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,10240,0.33554400338066953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,3072,0.07955644528071086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,2560,0.0122097780307134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,2560,0.06521955463621351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,8192,0.27286844783359104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,2048,0.010336000058386061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,2048,0.053195556004842125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,7168,0.24118667178683809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,1536,0.00851999968290329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,1536,0.04282666577233208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,6144,0.20957599745856392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,1024,0.006456888798210356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,1024,0.03128977616628011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,5120,0.17858666843838164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,768,0.005642666584915585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,768,0.02611911131276025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,4096,0.14571555455525717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,512,0.004836444639497333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,512,0.021856889128684998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,3584,0.1282151142756144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,256,0.004202666795916027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,256,0.017995556195576985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,2560,0.09593688779407078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,3072,0.11063644621107314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,128,0.003671110918124517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,128,0.01613333324591319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,64,0.0034675554682811103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,64,0.016020443704393175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,32,0.003540444291300244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,32,0.01574222246805827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,65536,0.11847200658586289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,65536,1.5578329298231337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,2048,0.08188533120685153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,16384,0.03784177700678507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,1536,0.06810666455162896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,16384,0.40560799174838597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,12288,0.03108355402946472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,1024,0.05270222160551283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,12288,0.30888978640238446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,10240,0.030433777305814955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,10240,0.26046755578782826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,768,0.04464088877042135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,8192,0.02236533330546485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,512,0.03769777880774604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,256,0.0322728885544671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,8192,0.21090488963656953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,7168,0.020051555501090158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,7168,0.18585511048634848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,6144,0.01607911123169793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,6144,0.16098754935794407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,128,0.029823111163245306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,5120,0.01401511165830824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,5120,0.13598755995432535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,12288,0.3897679911719428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,16384,0.509536001417372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,4096,0.011936888926559024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,65536,1.9518896738688152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,4096,0.10945244630177815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,3584,0.010321777727868822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,3584,0.09432622459199692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,3072,0.0090844440791342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,10240,0.33011823230319554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,3072,0.07772622505823772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,2560,0.007695111135641734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,2560,0.06378310918807983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,8192,0.26878754297892254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,2048,0.006627555522653792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,7168,0.23878844579060873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,2048,0.052383999029795326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,6144,0.20696711540222168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,1536,0.005903999838564131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,1536,0.042450666427612305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,1024,0.004966222163703707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,1024,0.029839111698998347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,5120,0.17587378289964464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,768,0.004518222063779831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,768,0.025702221526039973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,4096,0.14394133620791966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,512,0.00407644444041782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,512,0.02162222233083513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,3584,0.12584266397688124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,256,0.003640888879696528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,256,0.017916444275114272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,3072,0.10902400149239434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,128,0.0033724444607893624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,128,0.015925332903862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,2560,0.09411111142900254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,64,0.003149333306484752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,64,0.015511110424995422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,32,0.0032924444725116095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,32,0.015258666541841296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,65536,0.10848977830674912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,65536,1.5567297405666773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,1536,0.06600888570149739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,2048,0.08030222521887885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,16384,0.033088889386918806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,1024,0.05031911200947232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,16384,0.40473421414693195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,12288,0.02642311155796051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,12288,0.3084995481703016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,10240,0.022826666633288067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,10240,0.2601635456085205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,768,0.04275911053021749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,8192,0.014741332994567024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,8192,0.2111253341039022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,512,0.0369137790468004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,7168,0.013542222479979197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,7168,0.1867039998372396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,128,0.02966844373279148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,256,0.03141866789923774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,6144,0.015517334143320719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,6144,0.16109332773420545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,5120,0.01422755585776435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,5120,0.13556622134314641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,16384,0.5017795562744141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,12288,0.38364622328016496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,4096,0.010804444551467896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,4096,0.10898755656348334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,3584,0.009862222605281407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,65536,1.918984095255534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,3584,0.09420977698432074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,3072,0.00923111124171151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,3072,0.07777777645323011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,10240,0.3244275516933865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,2560,0.007273777491516537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,2560,0.06405866808361478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,8192,0.26437422964307994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,2048,0.0063946665161185795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,2048,0.052560001611709595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,7168,0.23399734497070312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,1536,0.005717333406209946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,6144,0.20334222581651476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,1536,0.04246044490072462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,5120,0.17333243952857125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,1024,0.004753777964247597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,1024,0.030412445465723675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,768,0.004492444296677907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,768,0.025941333836979334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,4096,0.1410177813635932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,512,0.003882666842805015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,512,0.021329777108298406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,3584,0.12399733066558838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,256,0.0035457776652442086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,256,0.01770755483044518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,3072,0.10751199722290039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,128,0.003216888962520493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,128,0.015988444288571674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,2560,0.09227999713685776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,64,0.0030986666679382324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,64,0.015229332778188916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,32,0.003177777760558658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,32,0.015045333239767285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,65536,0.10434933503468831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,65536,1.5567039913601344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,16384,0.032024890184402466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,16384,0.4043084515465631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,12288,0.02481777800454034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,2048,0.07882755332522921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,10240,0.02016622159216139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,12288,0.3081555631425646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,8192,0.012075555821259817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,10240,0.2598133352067736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,7168,0.010496888723638324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,8192,0.2107928858862983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,6144,0.01519555515713162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,7168,0.18607289261288115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,5120,0.009420444567998251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,6144,0.16117600599924722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,5120,0.13567999998728433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,4096,0.00868000007337994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,3584,0.007626666790909237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,4096,0.10923555824491714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,3072,0.007165333463086023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,3584,0.09403466516070896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,2560,0.006607111129495833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,3072,0.07800622118843926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,2048,0.005763555566469829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,2560,0.06383644209967719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,1536,0.005079111291302575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,2048,0.05261333452330696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,1024,0.004393777913517422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,1536,0.04201777776082357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,768,0.004042666819360522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,1024,0.030468444029490154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,512,0.0037271111375755737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,768,0.025407110651334126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,256,0.003346666693687439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,512,0.021418665846188862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,128,0.0030737777964936364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,256,0.01753866672515869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,64,0.00295733329322603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,128,0.01569422251648373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,32,0.002992000016901228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,64,0.015134221977657743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,32,0.015015999476114908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,65536,0.10102666748894586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,16384,0.0292595558696323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,16384,0.4045848846435547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,12288,0.023771555887328252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,12288,0.3081493377685547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,10240,0.017128888103697035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,65536,1.557112905714247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,8192,0.011010666688283285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,10240,0.259805334938897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,7168,0.01012711144155926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,8192,0.21091732713911268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,6144,0.009350222018029954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,7168,0.18573066923353407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,5120,0.00851733320289188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,6144,0.16123999489678278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,4096,0.007897777689827813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,5120,0.135279999838935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,3584,0.0075599998235702515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,4096,0.10928889115651448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,3072,0.007231999602582719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,3584,0.09396800067689683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,1024,0.04991288979848226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,2560,0.006243555496136348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,3072,0.07784799734751384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,1536,0.06503111124038696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,2048,0.0055440002017551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,2560,0.06401422288682726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,1536,0.00498933345079422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,2048,0.05232444405555725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,1024,0.004299555387761858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,1536,0.04194844431347317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,768,0.004015111260943943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,1024,0.030269334713617962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,512,0.003660444584157732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,768,0.025391111771265667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,256,0.003378666730390655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,768,0.042633778519100614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,512,0.02145599987771776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,128,0.003067555526892344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,256,0.01756533318095737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,64,0.0029271110478374693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,128,0.015975111060672335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,32,0.003006222140457895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,64,0.015236443943447538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,32,0.01520088811715444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,512,0.03637333379851447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,256,0.0310791101720598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,16384,1.8023732503255208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,12288,1.3766560024685328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,12288,1.3210168414645724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,16384,1.795248031616211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,12288,2.06020270453559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,10240,1.6838853624131944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,10240,1.1557351218329537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,8192,0.90903017255995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,10240,1.1079173617892795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,8192,1.3619964387681749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,7168,0.7920328776041666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,8192,0.8855111334058973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,128,0.029887109994888306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,7168,0.7742177645365397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,16384,2.716226577758789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,6144,0.6787599987453885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,6144,0.6804018020629883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,7168,1.245822270711263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,5120,0.5585253503587511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,5120,0.5575937694973415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,4096,0.44878578186035156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,5120,0.7932373152838813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,6144,1.0323013729519315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,4096,0.4651173485649957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,3584,0.4064951207902696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,3584,0.39117156134711367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,4096,0.6527768770853678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,3072,0.3335217899746365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,3072,0.50400177637736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,3072,0.3562328815460205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,2560,0.27739911609225804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,2048,0.22448711925082734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,2560,0.4251386589474148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,3584,0.5942808787027994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,2560,0.30183911323547363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,1536,0.1684115595287747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,2048,0.2557884322272407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,1536,0.20398578378889295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,2048,0.350491550233629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,1024,0.12166133191850449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,1024,0.18636088901095918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,1024,0.15651110808054605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,768,0.14929689301384821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,768,0.10249066352844238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,768,0.13310221831003824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,512,0.08743288781907822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,256,0.07621688975228204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,512,0.1074906653828091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,256,0.09100266959932114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,128,0.06321689155366686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,1536,0.2680133448706733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,128,0.07080444362428454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,256,0.07735377550125122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,64,0.05853244331147936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,512,0.112152894337972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,128,0.08869777785407172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,32,0.05712266763051351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,64,0.07260444429185656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,32,0.07138311200671725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,12288,0.4424453311496311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,16384,0.580126232571072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,16384,0.6124497519599067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,12288,0.4471457799275716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,10240,0.37567732069227433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,12288,0.4672062132093642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,10240,0.3736071056789822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,10240,0.39582665761311847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,16384,0.6124880048963759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,8192,0.2917368941836887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,7168,0.25229422251383465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,8192,0.3005982240041097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,8192,0.32079021135965985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,7168,0.26884977022806805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,7168,0.2807511223687066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,6144,0.21483111381530762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,6144,0.2443866729736328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,6144,0.23111022843254936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,5120,0.17615111668904623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,5120,0.20612622631920707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,5120,0.1960444450378418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,4096,0.15953511661953396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,3584,0.12565599547492132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,4096,0.14355644914839003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,4096,0.17277688450283477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,3584,0.14029867119259304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,3584,0.15543199910057917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,3072,0.10959466960695054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,3072,0.13736622863345677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,3072,0.12191555235120986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,2560,0.09432622459199692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,2560,0.10489333338207668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,2560,0.12028355068630642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,2048,0.07649777995215522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,2048,0.08585421906577216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,65536,2.5198790232340493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,2048,0.10638755559921265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,1536,0.08732799688975017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,1536,0.06220799684524536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,1536,0.06763822502560086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,65536,2.3237955305311417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,768,0.041448887851503156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,1024,0.07012710968653361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,768,0.04223377837075127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,512,0.031713777118259005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,1024,0.04979111088646782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,768,0.06199466519885593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,512,0.037017779217825994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,256,0.022675555613305833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,128,0.017456000049908955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,512,0.05411110983954536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,256,0.03127911024623447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,64,0.014926221635606555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,256,0.04706933432155185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,128,0.04339111182424757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,32,0.015551111764378019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,128,0.028734223710166082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,64,0.02959111001756456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,32,0.030564443932639226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,1024,0.04885244369506836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,16384,0.4746115472581651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,16384,0.520634651184082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,12288,0.3379066785176595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,12288,0.36313509941101074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,12288,0.4012426535288493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,16384,0.4432808823055691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,10240,0.3085093233320448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,10240,0.28361688719855416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,8192,0.2384240097469754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,10240,0.3372960090637207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,8192,0.2777022255791558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,8192,0.23054401079813638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,7168,0.20792088243696424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,7168,0.24392800860934785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,7168,0.2016915612750583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,6144,0.17701778146955702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,65536,1.89317872789171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,6144,0.2125635544459025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,6144,0.1737431155310737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,5120,0.14481511380937365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,5120,0.1763662232293023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,5120,0.14763200283050537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,4096,0.11903911166720921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,4096,0.1474293337927924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,3584,0.10428266392813788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,4096,0.12067200077904595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,3584,0.10790310965643989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,3072,0.09042755762736003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,3584,0.13449600007798937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,3072,0.12079377969106038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,3072,0.09331022368537055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,2560,0.07924889193640815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,2048,0.06493510802586873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,2560,0.10536444187164307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,2048,0.09043288893169826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,2560,0.08033066987991333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,65536,2.5997305976019964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,1536,0.05400977863205803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,1024,0.038929777012930974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,2048,0.0666560000843472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,1536,0.07695555686950684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,1024,0.0440577765305837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,1024,0.06279200315475464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,768,0.032362666394975446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,1536,0.052410665485594005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,768,0.03714044557677375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,768,0.05585600270165337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,512,0.025719111164410908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,512,0.03216088811556498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,512,0.048912889427608915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,256,0.01774666706720988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,256,0.027408889598316614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,128,0.01387733303838306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,256,0.042245331737730235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,64,0.012273777690198688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,128,0.02496622171666887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,32,0.012539555629094442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,128,0.03940533267127143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,64,0.023771555887328252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,32,0.024078221784697637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,65536,2.030184851752387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,16384,0.4242657820383708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,16384,0.37578490045335555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,12288,0.28508443302578396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,12288,0.32954488860236275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,10240,0.2753617763519287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,10240,0.23868976698981392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,65536,1.9619892968071833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,8192,0.1937706602944268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,8192,0.22438311576843262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,65536,1.6913742489284938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,7168,0.18653689490424263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,7168,0.17058666547139487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,6144,0.14719911416371664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,6144,0.15668533907996282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,65536,1.7928763495551214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,5120,0.12433777915106879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,5120,0.12919644514719644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,16384,0.5283208953009711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,4096,0.1009075575404697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,4096,0.10617066754235162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,12288,0.4022871123419867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,3584,0.09112355444166396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,3584,0.09424622191323175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,10240,0.34096977445814347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,3072,0.07936177651087443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,3072,0.08234044578340319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,8192,0.27920354737175834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,2560,0.06753510899013944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,2560,0.07154577970504761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,65536,2.0245422787136502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,7168,0.2440862125820584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,2048,0.05659288830227322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,2048,0.05932355589336819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,1536,0.04556266797913445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,1536,0.049588445160124034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,6144,0.20866310596466064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,1024,0.033062222931120135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,1024,0.03793066740036011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,5120,0.17919554975297716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,768,0.027043556173642475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,768,0.033478223615222506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,4096,0.1486604478624132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,512,0.01999022232161628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,512,0.029789331886503432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,3584,0.13568088743421766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,256,0.01421688828203413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,256,0.02535555594497257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,3072,0.12048889531029595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,128,0.011085333095656501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,2560,0.10650577810075547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,128,0.02273688879277971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,64,0.009360888765917884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,64,0.022563555174403723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,32,0.00945955514907837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,32,0.02205955485502879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,1536,0.07733066876729329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,2048,0.09238400061925252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,768,0.05546489026811388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,1024,0.06370044416851468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,16384,0.3782497776879205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,65536,1.4895057678222656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,16384,0.29705243640475804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,256,0.0440959996647305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,512,0.049159109592437744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,12288,0.29117955101860893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,16384,0.4331839879353841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,12288,0.333751122156779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,10240,0.24568711386786568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,65536,1.653633753458659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,12288,0.2230630980597602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,128,0.04012533360057407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,10240,0.18858133422003853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,65536,1.3704808553059895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,10240,0.28125156296624076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,8192,0.19194666544596353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,7168,0.16375199953715006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,8192,0.23183910051981607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,7168,0.20111378033955893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,6144,0.13826933172014025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,6144,0.11615555816226536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,6144,0.17407111326853433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,7168,0.13360622194078234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,8192,0.1525591082043118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,5120,0.09875822067260742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,5120,0.11596443918016221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,3584,0.08394400278727214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,5120,0.1485946708255344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,4096,0.09454400009579128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,3584,0.11477511458926731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,4096,0.1236577828725179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,3072,0.07508533530765109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,3072,0.10126044352849324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,4096,0.08095288938946195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,2560,0.06467288732528687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,2048,0.05292444427808126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,2560,0.0898853341738383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,3584,0.07492444250318739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,1536,0.03567466802067227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,2048,0.07838666439056396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,3072,0.06268266836802165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,1024,0.025769778423839148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,1536,0.045145776536729604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,2560,0.053730666637420654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,1024,0.03469689024819268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,1536,0.06739555464850532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,768,0.02102577851878272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,1024,0.05387022097905477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,2048,0.04452888833151924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,768,0.03106133474244012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,512,0.016369778248998854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,512,0.027375110321574744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,768,0.04786133435037401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,256,0.011961778004964193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,256,0.022638221581776936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,256,0.03709244396951463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,512,0.04315822323163351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,128,0.008926221893893348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,128,0.020111999577946134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,64,0.007651555869314406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,32,0.008002666963471307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,64,0.02069422271516588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,128,0.035078220897250705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,32,0.020227554771635268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,16384,0.37466756502787274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,16384,0.4398497740427653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,12288,0.22439644071790907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,16384,0.29642756779988605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,12288,0.28540534443325466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,12288,0.33628177642822266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,10240,0.18879911634657118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,10240,0.2395253446367052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,10240,0.28416532940334743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,8192,0.15107733673519558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,8192,0.1871591144137912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,7168,0.16148000293307835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,8192,0.231711122724745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,7168,0.13365066051483154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,6144,0.13923466205596924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,6144,0.17615911695692274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,7168,0.20263377825419107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,65536,1.4649510913425023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,6144,0.11620799700419109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,5120,0.11320000224643284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,4096,0.09223022063573201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,5120,0.14988711145189074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,4096,0.12447999583350287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,3584,0.08197866545783149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,5120,0.0973093310991923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,4096,0.07938310835096571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,3072,0.07199111249711779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,3584,0.11529955599043105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,3072,0.10123022397359212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,3584,0.0708364446957906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,2560,0.06367999977535672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,2048,0.044419556856155396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,2048,0.05255377623769972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,2560,0.09132089217503865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,1536,0.03496444556448195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,2048,0.07697511381573148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,1536,0.044890665345721774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,3072,0.06213866339789497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,1024,0.02514133354028066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,1536,0.06807111369238959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,2560,0.053132444620132446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,1024,0.03398666779200236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,768,0.02090488870938619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,1024,0.05560266971588135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,512,0.01608888970481025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,768,0.03048711021741231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,512,0.026991999811596338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,768,0.04907022251023186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,256,0.011677333050303988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,256,0.021879111727078755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,512,0.0443484452035692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,128,0.008791111409664154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,256,0.03802400165134006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,128,0.020205333828926086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,64,0.007440889047251807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,32,0.008067555725574493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,64,0.020592888196309406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,128,0.03524000114864773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,32,0.020015110572179157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,65536,1.6493093702528212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,16384,0.3304426670074463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,12288,0.25237777498033315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,65536,1.1998400158352323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,12288,0.1749795542822944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,16384,0.22672533988952637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,10240,0.21233688460456002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,10240,0.14485955238342285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,8192,0.17291822698381212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,8192,0.1170568863550822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,65536,1.290122667948405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,7168,0.10237955384784275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,7168,0.14490666654374865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,6144,0.088755554623074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,6144,0.12316889233059353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,65536,1.0218755933973525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,5120,0.07741689019733004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,5120,0.0995840032895406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,16384,0.39047911432054305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,12288,0.2988862196604411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,4096,0.06131644381417168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,4096,0.08093066347969903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,10240,0.25422578387790257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,3584,0.07309333483378093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,3584,0.05511022276348538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,8192,0.20803377363416883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,3072,0.06410311328040229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,3072,0.048207110828823514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,2560,0.04144000013669332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,2560,0.0576488905482822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,7168,0.1799386607276069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,2048,0.03417244553565979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,65536,1.4663546880086262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,2048,0.047169778082105845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,1536,0.027080888549486797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,1536,0.03923911187383864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,6144,0.15683733092414007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,1024,0.020232889387342665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,1024,0.031967110104031034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,5120,0.1342702176835802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,768,0.016944888565275405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,768,0.02847111225128174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,3584,0.10135644674301147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,512,0.013205333716339536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,512,0.024749333659807842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,4096,0.11150399843851726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,256,0.009393778112199571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,256,0.020062221421135795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,2560,0.08185155524147882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,128,0.007319110963079665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,128,0.01882844501071506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,768,0.04508799976772732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,64,0.006393777827421824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,64,0.01898755629857381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,32,0.00674311113026407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,32,0.018922666708628338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,3072,0.09149155351850723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,1536,0.05971466832690769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,2048,0.0708364446957906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,65536,1.259303092956543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,16384,0.1883280012342665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,1024,0.049867557154761426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,16384,0.3223022090064155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,12288,0.2472426626417372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,12288,0.14353511068556044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,65536,0.8418720033433703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,10240,0.12064533763461643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,10240,0.20910222000545928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,8192,0.0982453359497918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,8192,0.1639662186304728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,128,0.033237334754731916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,7168,0.0865137775739034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,256,0.0355404449833764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,7168,0.1419502231809828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,512,0.04050044549836053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,6144,0.07502044571770562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,6144,0.11998844146728516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,5120,0.06339110930760701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,5120,0.09838577773835923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,16384,0.37073598967658145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,65536,1.3827440473768446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,4096,0.0796497795316908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,4096,0.05167644553714328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,12288,0.2880480024549696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,3584,0.07099911239412096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,3584,0.04638044370545281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,3072,0.04063911239306132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,3072,0.06258755260043673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,10240,0.2428213225470649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,2560,0.03533688849873013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,2560,0.05592177973853218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,8192,0.19937333795759413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,2048,0.029719998439153034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,2048,0.04691822330156962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,7168,0.17503378126356336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,1536,0.023323555787404377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,1536,0.03882577684190538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,6144,0.15158932738833958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,1024,0.017441777719391715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,1024,0.031453334622912936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,5120,0.1290924416648017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,768,0.014721777704026965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,768,0.02795555525355869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,3584,0.09948266877068414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,512,0.011911110745535957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,512,0.024263999528355066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,4096,0.10896799961725871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,256,0.008614222208658854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,256,0.020624889267815482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,3072,0.08909244669808282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,128,0.0069822221994400024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,128,0.018759111563364666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,2560,0.08043733570310804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,64,0.006228444476922353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,64,0.018960888187090557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,32,0.00647555540005366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,32,0.018909333480728995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,2048,0.06909244590335421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,768,0.04463911056518555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,1536,0.057968888017866343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,65536,1.095478269788954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,16384,0.14935021930270725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,16384,0.28291111522250706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,16384,0.34632444381713867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,1024,0.04961333341068692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,65536,0.5991013314988878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,12288,0.1132871045006646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,10240,0.18164977762434217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,12288,0.21669422255622017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,12288,0.2646622127956814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,65536,1.2881440056694877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,10240,0.09652977519565159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,10240,0.22574400901794434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,8192,0.14592444896697998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,8192,0.18236445056067574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,8192,0.07790578073925443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,7168,0.1250986655553182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,7168,0.16127644644843206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,6144,0.10494222243626912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,7168,0.06916266679763794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,6144,0.14063555664486355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,5120,0.08612800306744045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,6144,0.05940622091293335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,5120,0.12012178368038601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,4096,0.06947822040981717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,3584,0.03801244497299194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,4096,0.10070221953921848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,3584,0.06277244620853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,5120,0.05091822147369385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,4096,0.04273510972658793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,3072,0.03317600157525804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,3584,0.0918382207552592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,3072,0.05595466825697157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,2560,0.02854222059249878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,2560,0.05004444387223986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,3072,0.08312622043821546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,2048,0.023456000619464453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,2560,0.07517066929075453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,2048,0.041116442945268415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,1536,0.019164444671736825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,1024,0.014581332604090372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,2048,0.06407288710276286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,1536,0.03485066692034403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,1024,0.028022223048739966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,768,0.012424888710180918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,1536,0.054508444335725575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,512,0.010088889135254754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,1024,0.0466542210843828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,768,0.02511288887924618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,512,0.02090222140153249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,256,0.007380444142553542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,768,0.04225422276390923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,512,0.03631110986073812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,256,0.018396443790859647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,128,0.006095110956165526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,256,0.03329688972897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,128,0.0173688895172543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,64,0.0053839998112784485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,128,0.03182044294145372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,32,0.005576888720194499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,64,0.017475555340449016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,32,0.017442666822009616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,16384,0.274071110619439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,16384,0.15242311689588758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,128,0.03271733389960395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,16384,0.3476248847113715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,12288,0.2127671109305488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,256,0.03692444496684604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,12288,0.11473600069681804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,10240,0.17795822355482313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,65536,1.0768053266737196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,10240,0.22555377748277453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,12288,0.2656124432881673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,10240,0.09626933601167466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,8192,0.1418328947491116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,7168,0.12300888697306316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,8192,0.18375910653008354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,7168,0.1616684463289049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,6144,0.10347200102276272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,8192,0.07885689205593534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,7168,0.06908266411887275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,6144,0.05581777625613742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,6144,0.1402453316582574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,5120,0.08450311422348022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,4096,0.03676977753639221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,5120,0.11982756190829807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,4096,0.06843022505442302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,5120,0.04965866605440775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,3584,0.032842665910720825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,4096,0.10119999779595269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,3072,0.03160355488459269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,3584,0.062182221147749156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,65536,1.2938071356879342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,3072,0.05482044484880236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,3584,0.09123644563886855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,2560,0.024998222788174946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,3072,0.08326844374338786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,2048,0.02092711130777995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,2560,0.0490862230459849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,2048,0.04057066639264425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,2560,0.07461866405275133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,1536,0.017253332667880587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,2048,0.06378755304548475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,1536,0.03415111038419936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,1024,0.01239999963177575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,1536,0.05532266696294149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,1024,0.027673777606752183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,768,0.010230222509966956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,1024,0.047084444099002414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,512,0.007984889050324758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,768,0.024479111035664875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,768,0.04169333312246534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,512,0.020907556017239887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,256,0.006210666563775804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,512,0.037159999211629234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,128,0.005295111073387994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,256,0.01847999956872728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,64,0.005015999906592899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,256,0.03382755650414361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,128,0.01738222274515364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,32,0.0053288886944452924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,64,0.017515555024147034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,128,0.03236888845761617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,32,0.01720977822939555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,512,0.04001599881384108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,65536,0.5825929111904568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,16384,0.2751893202463786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,16384,0.11512445078955756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,16384,0.32414311832851833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,12288,0.20993955930074057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,10240,0.1772488885455661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,12288,0.08690577745437622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,10240,0.21071910858154297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,12288,0.2484417756398519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,8192,0.14067910777197942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,8192,0.1711039940516154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,10240,0.07400000095367432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,8192,0.0597662197219001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,7168,0.12204710642496745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,7168,0.05335644549793667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,7168,0.14966044161054823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,6144,0.10100533564885457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,6144,0.1296186712053087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,5120,0.03960799839761522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,5120,0.08360355430179173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,65536,1.0651893615722656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,6144,0.04630755715899997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,4096,0.03275466627544827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,5120,0.11134755611419678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,4096,0.06856444146898058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,3584,0.02888622217708164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,4096,0.09339733256234063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,3584,0.061920000447167285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,3072,0.025608888930744592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,3072,0.05500533183415731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,2560,0.022287999590237934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,3584,0.08608711428112453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,2560,0.04914933443069458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,3072,0.07780800263086955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,2048,0.01870666609870063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,2560,0.07024533218807645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,2048,0.04058311051792569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,1536,0.015292444162898593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,2048,0.06013244390487671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,1536,0.03450222147835626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,1024,0.011667555405033959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,1024,0.027062222361564636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,1536,0.05365688933266533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,768,0.01016711112525728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,1024,0.04500000013245476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,512,0.007845333053006066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,768,0.023781332704755995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,512,0.021007110675175984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,768,0.040163556734720864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,256,0.006191111273235745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,512,0.0350488887892829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,256,0.018575110369258456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,128,0.005199999858935674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,65536,1.198822233412001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,128,0.01719377769364251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,64,0.004641777939266629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,256,0.032391998502943255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,32,0.004920000003443824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,64,0.017216889394654166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,128,0.031353778309292264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,32,0.017082666357358296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,16384,0.2726408905453152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,16384,0.31446488698323566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,16384,0.10042933622996013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,12288,0.07614666885799833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,12288,0.20752267042795816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,65536,0.4453626738654242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,10240,0.17551110850440133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,12288,0.24380087852478027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,10240,0.20663821697235107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,8192,0.1394017802344428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,10240,0.0649991101688809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,8192,0.16736443837483725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,7168,0.12025154961480035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,8192,0.053123556905322604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,6144,0.040962666273117065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,7168,0.04728177852100796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,7168,0.14759733941819933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,6144,0.10064088635974461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,5120,0.03535022338231405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,5120,0.08301511075761583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,6144,0.12751111719343397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,4096,0.029294222593307495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,5120,0.10937066872914632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,4096,0.06794933478037517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,3584,0.026360000173250835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,4096,0.0934000015258789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,3072,0.023278221487998962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,3584,0.06172888808780246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,3584,0.08567821979522705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,2560,0.020233778489960563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,3072,0.05463111069467333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,3072,0.07750666803783841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,65536,1.051579581366645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,2048,0.01753866672515869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,2560,0.048531555467181735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,2560,0.0709128909640842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,65536,1.1582302517361112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,1536,0.014280888769361707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,2048,0.040060444010628596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,1536,0.03417155477735732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,1024,0.02633511192268795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,2048,0.05912977457046509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,1024,0.011138666835096149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,1536,0.05342400074005127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,768,0.009671111073758867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,1024,0.04352088769276937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,768,0.02327288852797614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,512,0.007669332954618666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,768,0.039649777942233615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,512,0.020557334025700886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,256,0.005984000033802456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,256,0.018304889400800068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,512,0.03427911135885451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,128,0.005061333378156026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,64,0.0046746668716271715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,256,0.0321777794096205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,128,0.017125333348910015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,65536,0.3876604504055447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,128,0.030682666434182063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,32,0.004770666774776247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,64,0.017102221647898357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,32,0.017292444904645283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,12288,0.18305244710710314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,16384,0.2356568972269694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,12288,0.059191114372677274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,16384,0.07756355735990736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,10240,0.15363289250267878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,10240,0.05007822314898173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,8192,0.12192622820536296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,8192,0.04081777731577555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,7168,0.036539554595947266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,7168,0.104839113023546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,65536,0.9009555180867513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,6144,0.031649778286616005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,6144,0.08778222401936848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,65536,0.30745689074198407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,5120,0.02709688908523983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,5120,0.07187910874684651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,12288,0.2370097902086046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,4096,0.022890667120615642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,16384,0.303147554397583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,4096,0.060361776087019176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,3584,0.020631111330456205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,3584,0.054530666934119336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,10240,0.19914755556318495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,3072,0.018000000052981906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,3072,0.04741866721047295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,8192,0.1608533329433865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,2560,0.015803555647532146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,2560,0.042768889003329806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,7168,0.14088443915049234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,2048,0.013567110730542077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,2048,0.03557422094874912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,65536,1.1276266309950087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,6144,0.12192711565229629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,1536,0.01131377783086565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,1536,0.03087288803524441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,1024,0.008773333496517604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,1024,0.023680888944202002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,5120,0.10471910900539821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,768,0.007258666886223688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,768,0.02129688858985901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,4096,0.08856355481677586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,512,0.006163555714819167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,512,0.01906133360332913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,3584,0.08210221926371257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,256,0.005079999979999331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,256,0.01732711162832048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,3072,0.0745004415512085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,128,0.004426666845877965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,2560,0.0669271084997389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,128,0.0158933334880405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,64,0.003996444245179494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,2048,0.05665333403481377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,64,0.01590666671593984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,32,0.004231111043029361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,32,0.01573244399494595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,768,0.03684799869855245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,1536,0.05019466744528877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,1024,0.042282667424943715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,16384,0.05910755528344048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,16384,0.2309244473775228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,65536,0.8709920247395834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,65536,0.23438223203023276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,12288,0.1762524445851644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,12288,0.047616001632478505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,10240,0.04862311151292589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,10240,0.14866756068335638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,8192,0.03168088859981961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,8192,0.11920444170633952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,256,0.03195377853181627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,128,0.030074665943781536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,512,0.03377155462900797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,7168,0.030793779426150854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,7168,0.10270044538709854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,6144,0.03128977616628011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,6144,0.08575466606352065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,5120,0.023183999790085688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,5120,0.07077333662245008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,16384,0.3015315532684326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,4096,0.017584888471497428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,12288,0.23397689395480684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,4096,0.05815822548336453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,10240,0.1975946691301134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,3584,0.015987555185953777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,3584,0.05323822299639384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,3072,0.014368888404634265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,3072,0.04681422313054403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,65536,1.1188248528374565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,8192,0.15980888737572566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,2560,0.012697777814335294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,2560,0.041788445578681103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,2048,0.010798222488827176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,2048,0.03481066558096144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,7168,0.14031378428141275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,1536,0.009082666701740688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,1536,0.02994577752219306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,6144,0.12132978439331055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,1024,0.007231111327807109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,1024,0.02311555544535319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,5120,0.1037644412782457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,768,0.006233777850866318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,768,0.02089333368672265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,3584,0.08140977885988024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,512,0.0053031109273433685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,512,0.018775999546051025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,2560,0.06691910823186238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,256,0.004556444370084339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,256,0.016912889149453905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,4096,0.08898666832182144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,128,0.004042666819360522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,128,0.015829333000712924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,3072,0.07461244530147977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,64,0.003681777666012446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,64,0.015743111570676167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,32,0.0038364442686239877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,2048,0.056622220410241016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,32,0.015599111715952555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,65536,0.8768622080485026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,65536,0.16084089544084337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,16384,0.049066666099760264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,16384,0.23162489467196992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,768,0.03666933377583822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,12288,0.04276355438762241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,12288,0.1758586698108249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,1536,0.049739556180106274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,10240,0.04266311062706841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,10240,0.14885244104597303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,1024,0.042266666889190674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,8192,0.032280001375410296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,8192,0.1174444490008884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,128,0.029642668035295274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,7168,0.029232889413833618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,7168,0.10371555884679158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,256,0.03103999959097968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,6144,0.02603200078010559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,6144,0.08607644504970974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,512,0.033067554235458374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,5120,0.022522666388087805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,5120,0.07030044661627875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,65536,1.0659866333007812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,16384,0.2879066732194689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,4096,0.016736000776290894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,4096,0.05776888794369168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,3584,0.015983111328548856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,3584,0.052654223309622876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,12288,0.22191288736131456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,3072,0.014010666145218743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,3072,0.046668443414900035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,10240,0.18865956200493705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,2560,0.011754666765530905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,2560,0.04068177938461304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,8192,0.1540622181362576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,7168,0.13436888323889837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,2048,0.010090666512648264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,2048,0.03477333320511712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,1536,0.008095111283991072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,1536,0.029418667157491047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,6144,0.11582667297787136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,1024,0.006245333287451003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,1024,0.023073777556419373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,5120,0.10050400098164876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,768,0.005560889012283749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,768,0.020896888441509668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,4096,0.08534666564729478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,3584,0.07869066794713338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,512,0.004808889081080754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,512,0.01923377811908722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,256,0.0041351111398802865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,256,0.016846223009957206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,3072,0.0711395541826884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,128,0.0036684444381131064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,128,0.015651555524932016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,2560,0.06398489077885945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,64,0.0034053333931499054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,64,0.015616888801256815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,32,0.003491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,32,0.01538044379817115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,65536,0.12403021918402778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,65536,0.8601937823825412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,2048,0.05553955501980252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,16384,0.03976088762283325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,16384,0.2290711137983534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,1536,0.04826133449872335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,12288,0.041146665811538696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,1024,0.03874489002757602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,12288,0.17521954907311332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,10240,0.03613777624236213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,10240,0.1474284463458591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,768,0.03555288910865784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,8192,0.026362667481104534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,8192,0.11814933353000218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,256,0.030631999174753826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,7168,0.02197155521975623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,7168,0.10248711374070908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,512,0.03234044379658169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,6144,0.02162044412559933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,6144,0.08492000235451592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,128,0.02924355533387926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,5120,0.017294221454196505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,5120,0.07010044654210408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,65536,1.0687991248236763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,4096,0.014436443646748861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,4096,0.05728889173931546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,16384,0.2899831136067708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,3584,0.013224889006879596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,3584,0.05227999885876974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,12288,0.22252000702752006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,3072,0.01237777786122428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,3072,0.04618488748868307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,10240,0.188481781217787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,2560,0.009834667046864828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,2560,0.04042133357789781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,8192,0.15343199835883245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,2048,0.00868266655339135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,7168,0.13423200448354086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,2048,0.034212443563673235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,1536,0.006807111203670502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,1536,0.028728889094458684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,6144,0.1163288884692722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,1024,0.005681777579916849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,1024,0.022820444570647344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,5120,0.09957689046859741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,768,0.004931555440028508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,768,0.020638222495714825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,4096,0.08452355861663818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,512,0.004270222038030624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,512,0.01864088906182183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,3584,0.07803199688593547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,256,0.0038088887102074097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,256,0.016515556308958266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,2560,0.06285866763856676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,128,0.0034097778714365433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,128,0.015618667006492615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,3072,0.07038222418891059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,64,0.0032008888406885993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,64,0.015580443872345818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,32,0.003342222215400802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,32,0.01548266741964552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,768,0.03497688968976339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,1536,0.04712444543838501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,2048,0.05429155627886454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,65536,0.09410044219758774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,16384,0.032700445916917585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,16384,0.2294551001654731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,12288,0.032871110571755305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,65536,0.8618888854980469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,12288,0.17504710621303984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,10240,0.029680887858072918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,1024,0.038105779223971896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,10240,0.1474684476852417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,8192,0.01907466683122847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,8192,0.11787199974060059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,128,0.028943998946083918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,7168,0.020206222931543987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,512,0.03219644559754266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,256,0.030985779232449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,7168,0.10242666800816853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,6144,0.01827822294500139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,6144,0.08454044659932454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,5120,0.016206221448050607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,5120,0.06966844532224867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,16384,0.27932000160217285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,12288,0.21541421943240693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,4096,0.011845333708657159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,4096,0.05747289127773709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,10240,0.18388266033596465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,65536,1.0340586768256295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,3584,0.011104889214038849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,3584,0.05223200056287977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,3072,0.009003555609120263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,3072,0.045886221859190196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,2560,0.007975111405054728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,2560,0.04014666544066535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,8192,0.1496595541636149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,2048,0.006829333388143116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,2048,0.03408177693684896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,7168,0.13075733184814453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,1536,0.0058853332367208265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,1536,0.02757511039574941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,6144,0.11354933844672309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,1024,0.005134222408135732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,1024,0.023157333334287006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,5120,0.09707822402318318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,768,0.004513777792453766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,768,0.020584889584117465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,3584,0.07563200261857775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,4096,0.08272622028986613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,512,0.004022222426202563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,3072,0.06859199868308173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,512,0.0185244447655148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,256,0.0035991109907627106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,256,0.016387555334303115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,128,0.0033582221302721235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,2560,0.0615288880136278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,128,0.015408888459205627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,64,0.0031831111345026228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,64,0.015377778146002026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,32,0.0033039999090962936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,32,0.015172445111804538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,65536,0.06971555285983615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,65536,0.8601199785868326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,2048,0.052946666876475014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,16384,0.023747555083698694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,16384,0.22927644517686632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,1536,0.045329779386520386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,12288,0.019268444842762418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,1024,0.038035555018319026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,768,0.03412444392840067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,12288,0.17366577519310844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,10240,0.01864177816443973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,10240,0.1456835534837511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,8192,0.01371911085314221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,8192,0.11750133832295735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,128,0.029126223590638902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,7168,0.013195555243227216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,256,0.029538667864269678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,7168,0.10229421986473931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,512,0.031537777847713895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,6144,0.013977777626779346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,6144,0.08474133412043254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,5120,0.012464000119103326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,5120,0.0693368911743164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,65536,1.0083786646525066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,4096,0.009018667042255402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,4096,0.056913779841529004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,16384,0.2727866702609592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,3584,0.00849333322710461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,3584,0.0517884459760454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,10240,0.1785502168867323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,3072,0.008160889148712158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,12288,0.20994044674767387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,3072,0.04576177729500664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,2560,0.007349333001507654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,2560,0.03997066617012024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,8192,0.1460097763273451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,2048,0.00674044465025266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,2048,0.03245688809288873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,6144,0.11026044686635335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,7168,0.12738666269514295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,1536,0.005850666513045629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,1536,0.027387556102540758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,1024,0.004899555610285865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,1024,0.02241955531968011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,5120,0.09432799948586358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,768,0.004546666724814309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,768,0.020287111401557922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,4096,0.08059466547436185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,512,0.003973333372010125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,512,0.018398221996095445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,3584,0.07428533501095243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,256,0.0035253332720862497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,256,0.0162471118900511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,3072,0.06720711125267877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,128,0.0032622222271230486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,128,0.015678221980730694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,2560,0.05980177720387777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,64,0.003056888985964987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,2048,0.05133066574732462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,64,0.01516177753607432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,32,0.0031111112071408164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,32,0.01478844384352366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,65536,0.05910933017730713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,1536,0.04346933298640781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,16384,0.016044444508022733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,1024,0.03711288836267259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,16384,0.228836456934611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,65536,0.8614533212449816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,12288,0.012175111307038201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,768,0.033766223324669734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,10240,0.011152000062995486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,12288,0.17493777804904512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,10240,0.14723911550309923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,8192,0.008946666287051307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,8192,0.11785777409871419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,512,0.03152889013290405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,7168,0.01253600004646513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,7168,0.10208977593315972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,256,0.029740442832310993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,6144,0.008918222453859117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,6144,0.0849191082848443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,128,0.029027554723951552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,5120,0.009332444104883406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,5120,0.06917866733339098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,16384,0.271941343943278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,65536,1.0050124062432182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,4096,0.008486222061845992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,12288,0.20906399355994332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,10240,0.1781111160914103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,4096,0.057215107811821826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,3584,0.00851288851764467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,3584,0.05172888769043816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,3072,0.007979555262459649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,3072,0.045532445112864174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,2560,0.006594666590293248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,8192,0.14470488495296902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,2560,0.03970311085383097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,2048,0.0058293334311909145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,2048,0.03240977724393209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,7168,0.127146667904324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,1536,0.005078222188684675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,1536,0.027365333504146997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,6144,0.10966133409076267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,1024,0.004407111141416761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,1024,0.022495999932289124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,5120,0.0947457750638326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,768,0.004048000193304486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,768,0.020255999432669747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,4096,0.07984889215893216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,3584,0.07340621948242188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,512,0.0036915557252036203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,512,0.018183110488785636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,256,0.0033679999825027254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,3072,0.06631466415193346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,256,0.016139555308553908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,128,0.00309599998096625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,128,0.0151591118839052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,2560,0.05922310882144504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,64,0.002942222274012036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,2048,0.05114577876196968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,32,0.003032888803217146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,64,0.014910222755538093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,32,0.014840000205569796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,65536,0.0588666664229499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,16384,0.012425777812798819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,16384,0.2291031148698595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,12288,0.010240000155236986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,65536,0.8611511124504937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,10240,0.008814222282833522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,12288,0.17465333143870035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,8192,0.007916444705592262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,10240,0.14725244045257568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,7168,0.007624000310897827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,8192,0.11769778198666042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,6144,0.009133333133326637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,7168,0.10208088821834987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,5120,0.008447999755541483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,6144,0.08441866768731011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,4096,0.007714666426181793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,1536,0.0446995562977261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,5120,0.06941777467727661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,3584,0.007520889242490132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,4096,0.05663910839292738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,3072,0.007149332927332983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,3584,0.05154577891031901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,2560,0.006362666686375936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,3072,0.04494666722085741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,2048,0.00564533347884814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,2560,0.03959200117323134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,1536,0.004986666556861666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,2048,0.03163555595609877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,1024,0.004299555387761858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,1536,0.027092445227834914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,768,0.003956444561481476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,1024,0.022310222188631695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,512,0.0036071112586392295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,768,0.02038666605949402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,256,0.0032684444967243406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,512,0.01813155578242408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,128,0.003031111011902491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,256,0.01626666718059116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,64,0.003040888864133093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,128,0.015048889650238885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,32,0.0029724445194005966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,64,0.015139554937680563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,32,0.014879999889267815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,65536,0.05674666828579373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,16384,0.011358222199810876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,16384,0.22849867078993055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,12288,0.009649777577983009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,65536,0.8591039975484213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,10240,0.00886311133702596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,12288,0.17431467109256318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,8192,0.007869333028793335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,10240,0.14716088771820068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,8192,0.118120895491706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,7168,0.0076248885856734375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,6144,0.0069875551594628235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,7168,0.10228621959686279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,5120,0.007358222372002072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,6144,0.0847475528717041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,4096,0.006879999819729064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,5120,0.06892889075809054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,3584,0.007291555404663086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,4096,0.05657511287265354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,3072,0.0069271110826068455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,3584,0.05169422096676297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,2560,0.006227555374304454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,3072,0.0449795557392968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,2048,0.00554044461912579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,2560,0.03840800126393636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,1536,0.004869333157936732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,2048,0.03233511249224345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,1024,0.004206222378545337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,1536,0.027222222752041284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,768,0.003863111138343811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,1024,0.022437334060668945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,512,0.0035084444615576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,768,0.020468443632125854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,256,0.0031591111587153543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,512,0.01822933389080895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,1024,0.03685422076119317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,128,0.002982222164670626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,256,0.01623288955953386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,128,0.015079110860824585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,64,0.002867555452717675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,32,0.0028684443483750024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,64,0.014866666661368476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,32,0.01479111115137736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,768,0.03430577781465318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,512,0.03127022253142463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,256,0.029489778810077246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,16384,1.3630879720052083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,16384,1.3989333046807184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,12288,1.0459253523084853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,12288,1.5113022062513564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,12288,1.0301840040418837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,10240,1.273313734266493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,10240,0.86880890528361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,8192,0.6782319810655383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,8192,0.6745448642306857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,10240,0.8767653571234809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,16384,1.9759065839979384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,7168,0.5955351193745931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,7168,0.5917057991027832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,6144,0.5130595631069607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,8192,1.0046763949924045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,7168,0.8232657644483777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,128,0.028648889727062646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,6144,0.7266071107652453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,5120,0.42191823323567706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,5120,0.443470213148329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,4096,0.34742400381300187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,6144,0.5300328996446397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,3584,0.30234845479329425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,4096,0.3564266628689236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,3584,0.323562675052219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,4096,0.4835235807630751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,3072,0.2589786582522922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,3584,0.41764089796278214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,5120,0.5906302134195963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,3072,0.2766595681508382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,3072,0.3605982197655572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,2560,0.2139208846622043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,2048,0.1749537785847982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,2048,0.20657955275641549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,1536,0.13708621925777859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,2560,0.30744179089864093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,1536,0.19457156128353545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,2560,0.2459182209438748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,1536,0.16593421830071345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,1024,0.13803199927012125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,768,0.08660088645087348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,1024,0.10111822022332086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,1024,0.1238586637708876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,512,0.07351733578575982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,2048,0.25151112344529897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,768,0.10891377925872803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,256,0.06249511241912842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,512,0.08311111397213407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,128,0.0502417782942454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,256,0.05765777826309204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,256,0.08162133561240302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,128,0.05790311098098755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,64,0.04495288928349813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,768,0.10975288682513767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,32,0.04543911086188423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,128,0.0761937763955858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,64,0.05990044275919596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,512,0.08921155664655898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,32,0.06159377760357327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,16384,0.43062133259243435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,12288,0.3317920102013482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,12288,0.3367573420206706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,16384,0.47044531504313153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,12288,0.35837599966261124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,16384,0.441814210679796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,10240,0.27856355243259007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,10240,0.30034489101833767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,10240,0.28271643320719403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,8192,0.21141333050198027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,8192,0.24708265728420684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,8192,0.22838311725192598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,7168,0.18567821714613172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,7168,0.22147288587358263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,6144,0.16115910477108425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,6144,0.19208710723453096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,6144,0.1742933326297336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,7168,0.20385332902272543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,5120,0.13453243838416204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,5120,0.1476915544933743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,5120,0.1668391095267402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,4096,0.11187999778323704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,4096,0.12000977993011475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,4096,0.13940000534057617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,3584,0.09900977876451279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,3584,0.1075013346142239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,3584,0.12709866629706487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,3072,0.08680178059471978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,3072,0.09335022502475315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,3072,0.11266933547125922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,2560,0.07390489180882771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,2560,0.09885955519146389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,2560,0.08350844515694512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,2048,0.06241155332989163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,1536,0.05279466509819031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,2048,0.0663964417245653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,1536,0.07469599776797824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,2048,0.08647733264499241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,1536,0.05252977874543932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,1024,0.04118577639261881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,768,0.032815112007988825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,768,0.03663644525739882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,1024,0.05963999695248074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,768,0.05524177683724297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,512,0.025598223010698955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,1024,0.039213332864973284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,256,0.01777688827779558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,512,0.03184977836079068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,512,0.048100445005628795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,256,0.027051554785834417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,128,0.013907555076811047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,256,0.04346488912900289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,128,0.024849777420361836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,64,0.012261333564917246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,128,0.0406915545463562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,64,0.024805333879258897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,32,0.012794666820102267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,65536,1.7238915761311848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,32,0.024421332610978022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,16384,0.3755902184380426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,65536,1.8760657840304906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,16384,0.3334711127811008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,12288,0.2897324562072754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,12288,0.25021245744493276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,10240,0.24186134338378906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,65536,1.5
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,10240,0.21187021997239855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,8192,0.18462222152286104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,8192,0.17159644762674967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,7168,0.1507706642150879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,65536,1.9905192057291667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,7168,0.1593493355645074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,6144,0.13037421968248156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,6144,0.13925777541266546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,65536,1.4646480348375108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,5120,0.11012266741858588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,5120,0.12028000089857314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,16384,0.3966960112253825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,4096,0.08896978033913507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,4096,0.09578488932715522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,12288,0.3100533220503065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,3584,0.07904533545176189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,65536,1.556653340657552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,10240,0.2576222154829237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,3584,0.08661155568228827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,3072,0.06891110870573255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,3072,0.07542666461732653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,2560,0.0589688883887397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,2560,0.06521155436833699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,8192,0.20909867021772596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,2048,0.04888888862397936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,2048,0.05551377932230631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,7168,0.18621955977545845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,1536,0.03913955556021796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,1536,0.047509332497914634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,6144,0.16290577252705893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,1024,0.028743111424975928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,1024,0.03677599959903293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,5120,0.14003111256493464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,768,0.02341777748531765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,768,0.03192889028125339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,4096,0.11866754955715603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,512,0.018384888768196106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,512,0.028018666638268366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,3584,0.10918755663765801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,256,0.013247999880048962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,256,0.02456444501876831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,2560,0.08698933654361302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,128,0.010071111222108206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,128,0.02169422143035465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,3072,0.09674577580557929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,64,0.00851733320289188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,64,0.02272266646226247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,32,0.008779555559158325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,32,0.021935110290845234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,2048,0.07565422190560235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,1536,0.06734310918384127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,1024,0.05337422092755636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,768,0.050336887439092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,16384,0.2973182201385498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,16384,0.32554488711886936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,256,0.0410444438457489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,128,0.03779200050565932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,65536,1.2910399966769748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,512,0.04439111219512092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,12288,0.24972711669074166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,10240,0.21030222045050728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,12288,0.22283734215630424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,10240,0.18840977880689833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,8192,0.1590026617050171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,8192,0.1512826681137085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,7168,0.1393431160185072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,7168,0.13334400124020046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,65536,1.2493137783474393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,6144,0.12526488304138184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,6144,0.11544177267286514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,5120,0.09742666615380181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,5120,0.10176177819569905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,16384,0.37718666924370664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,12288,0.29180177052815753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,4096,0.07954133219189115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,4096,0.08443644311692979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,10240,0.2442079914940728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,3584,0.07071022192637126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,65536,1.4189120398627386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,3584,0.0763733320766025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,3072,0.06788266367382474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,3072,0.0625964469379849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,2560,0.053708444039026894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,2560,0.05807199743058947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,8192,0.1975680059856839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,2048,0.04461866617202759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,2048,0.050092445479498975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,7168,0.17542755603790283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,1536,0.035036444664001465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,1536,0.0432222220632765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,6144,0.15444444285498724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,1024,0.025913778278562758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,1024,0.03350933392842611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,5120,0.13344621658325195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,768,0.020909332566791113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,768,0.029647111892700195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,3584,0.1041182213359409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,512,0.016166221764352586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,512,0.025811556312772963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,4096,0.11397066381242539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,256,0.011775111158688864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,256,0.022465777066018846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,2560,0.0818471113840739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,128,0.010592888626787396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,128,0.020971554848882887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,3072,0.0925262239244249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,64,0.00959644466638565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,64,0.020450666546821594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,32,0.010176888770527309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,32,0.020198222663667467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,1536,0.06387555599212646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,2048,0.07212799787521362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,768,0.047671112749311656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,1024,0.05084355672200521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,65536,1.1125546561347115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,16384,0.28378934330410427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,16384,0.22502401140001085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,12288,0.17049245039621988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,512,0.04234755701488919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,65536,1.00037776099311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,256,0.040067556831571795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,128,0.03625066743956672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,10240,0.18208444118499756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,12288,0.21588355965084502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,10240,0.1433137787712945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,8192,0.13813155227237278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,8192,0.11697155899471706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,7168,0.12109155125088161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,7168,0.10232533348931207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,6144,0.10523466931449042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,6144,0.08930310938093398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,5120,0.07567644119262695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,5120,0.08894844187630548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,16384,0.33255110846625435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,4096,0.06138844622506035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,4096,0.07266577747133043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,65536,1.2567822138468425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,3584,0.05552355448404948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,3584,0.06588888830608793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,10240,0.2151813374625312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,8192,0.17449688911437988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,12288,0.2548542287614611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,3072,0.060080885887145996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,3072,0.04843733376926846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,2560,0.050632887416415744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,2560,0.04113511244455973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,2048,0.03400711218516032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,2048,0.043113778034845986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,6144,0.13752888308631048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,1536,0.02700622214211358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,1536,0.03852444556024339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,7168,0.15961244371202257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,1024,0.020092444287406076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,1024,0.030419554975297716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,5120,0.1188853316836887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,768,0.016705777910020616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,768,0.026776888304286536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,3584,0.09280977646509807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,512,0.013252444565296173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,512,0.022856889499558344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,2560,0.07380088832643297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,256,0.009576000273227692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,256,0.019647111495335896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,4096,0.10041599803500706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,128,0.007526222202512953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,128,0.018559111489189994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,768,0.04526666800181071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,64,0.006638222270541721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,64,0.01906222270594703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,32,0.00684533350997501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,32,0.01867111192809211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,3072,0.08347466919157241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,1536,0.05942666530609131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,2048,0.06445955567889743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,1024,0.04764266808827718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,16384,0.19666666454739043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,16384,0.2768559985690647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,65536,1.0899795956081815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,12288,0.1504533290863037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,12288,0.21375021669599745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,65536,0.8085218005710177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,10240,0.127893328666687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,10240,0.17908178435431588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,8192,0.1035919984181722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,128,0.03406577640109592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,8192,0.1354586680730184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,256,0.037400000625186496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,7168,0.0914728906419542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,7168,0.11810221936967637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,6144,0.07914400100708008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,512,0.040672000911500715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,6144,0.10344266891479492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,5120,0.06807288858625624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,5120,0.08671644661161636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,16384,0.3275902271270752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,65536,1.2405964533487956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,4096,0.07078399923112658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,4096,0.05473600162400139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,3584,0.04977155394024319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,3584,0.06482755475574069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,12288,0.2510853343539768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,3072,0.0428746673795912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,3072,0.05902577771080864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,10240,0.21022844314575195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,8192,0.17050845093197295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,2560,0.04965600040223864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,2560,0.03678133421474033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,2048,0.030282666285832722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,2048,0.04236533244450887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,7168,0.15320889155069986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,1536,0.024307555622524683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,1536,0.03767911261982388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,6144,0.13422489166259766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,1024,0.017906667457686532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,1024,0.029955555995305378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,5120,0.11605511771308051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,768,0.014920888675583733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,768,0.026692445079485577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,3584,0.09063733286327785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,512,0.011385777758227454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,512,0.02293511066171858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,4096,0.09951021936204697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,256,0.008967111508051554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,256,0.019592000378502738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,2560,0.07274133629269071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,128,0.007739555504586961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,128,0.018558222386572096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,3072,0.08198666572570801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,64,0.0073831114504072405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,64,0.019222221440739103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,32,0.007623111208279927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,32,0.01885688967174954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,1536,0.06034666962093777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,2048,0.06409866942299737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,768,0.04465422365400526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,1024,0.046967999802695386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,65536,1.0771208869086373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,16384,0.16861955324808756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,16384,0.274254216088189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,12288,0.1280613342920939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,12288,0.2086177799436781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,10240,0.10776888661914402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,10240,0.17427289485931396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,65536,0.6959706412421333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,128,0.03391199972894456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,8192,0.08763200044631958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,256,0.036862221029069685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,8192,0.13234844472673205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,7168,0.11595733960469563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,7168,0.07718311415778266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,6144,0.06658844153086345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,512,0.0402746664153205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,6144,0.10065244303809272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,5120,0.056573331356048584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,5120,0.08481955528259277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,16384,0.3343279891543918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,4096,0.046513779295815356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,4096,0.06998311148749457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,65536,1.2459182739257812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,12288,0.25548267364501953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,3584,0.06345778041415744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,3584,0.042176889048682324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,3072,0.03695822093221877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,3072,0.05862400266859266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,10240,0.2141786681281196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,2560,0.031531555785073176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,2560,0.04865511258443197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,8192,0.1732177734375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,2048,0.026141333911154006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,2048,0.04158577654096816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,7168,0.15255022048950195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,1536,0.021124444074100916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,1536,0.035912000470691256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,6144,0.13382221592797175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,1024,0.016024889217482675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,1024,0.029380444023344252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,5120,0.1171991162829929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,768,0.013415111435784234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,768,0.026188444760110643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,4096,0.0990826686223348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,512,0.01073066641887029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,512,0.02295644415749444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,3584,0.09112444188859727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,256,0.007808888951937358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,256,0.019359111785888672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,3072,0.08309866984685262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,128,0.006488889041874144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,128,0.018497778309716117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,2560,0.07321155733532377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,64,0.005712888720962737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,64,0.018906666172875296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,32,0.00601244428091579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,32,0.018684445155991446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,2048,0.06385955545637342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,1536,0.056306666798061795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,768,0.04495911134613884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,1024,0.04708977871470981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,16384,0.14992800023820665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,16384,0.23053156005011666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,65536,0.9117333094278971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,12288,0.17831644746992323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,12288,0.11423733499315049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,512,0.03978044456905789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,10240,0.09641955296198527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,65536,0.5914604398939345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,10240,0.1499484380086263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,256,0.03570222192340427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,8192,0.1115422248840332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,8192,0.077565332253774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,7168,0.0693697796927558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,128,0.03421688742107815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,7168,0.09835644563039143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,6144,0.08598133590486313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,6144,0.06061422162585788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,5120,0.05143466591835022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,5120,0.07323200172848172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,16384,0.28912801212734646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,4096,0.0416515568892161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,4096,0.06013066901101006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,65536,1.0556364059448242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,3584,0.037215110328462385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,3584,0.05462133222156101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,12288,0.22002932760450575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,3072,0.03272799981964959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,3072,0.05120533373620775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,10240,0.1855377753575643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,2560,0.028447998894585505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,2560,0.04310311211480034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,8192,0.15192800097995335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,2048,0.02397511071628994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,2048,0.03665333324008518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,7168,0.13551555739508733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,1536,0.019073777728610568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,1536,0.03196355700492859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,6144,0.11914843983120388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,1024,0.01457244488928053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,1024,0.026363554928037856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,5120,0.10390310817294651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,768,0.012383999923865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,768,0.023864888482623633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,3584,0.08225333028369479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,512,0.009878222313192155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,512,0.020583111378881667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,4096,0.0892382197909885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,256,0.007288888924651676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,256,0.017994667092959087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,2560,0.06627555688222249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,128,0.00609777785009808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,128,0.017139555679427255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,1536,0.05220533410708109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,64,0.0053119998839166425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,64,0.01734044485621982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,32,0.005433777968088786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,32,0.01703288820054796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,3072,0.07492622401979235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,2048,0.057320886188083224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,768,0.04219466778967115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,65536,0.8784346580505371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,16384,0.22672621409098306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,1024,0.04383822282155355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,16384,0.11610044373406304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,65536,0.4452960226270888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,12288,0.1720142232047187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,12288,0.08843555715348984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,10240,0.07404977745480008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,10240,0.1453404426574707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,8192,0.060714668697781034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,8192,0.11057600047853257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,512,0.03772888912094964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,128,0.033315555916892156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,7168,0.09682133462693955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,256,0.03438399897681342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,7168,0.0533537765343984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,6144,0.08421244223912557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,6144,0.046105778879589505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,5120,0.03964711229006449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,5120,0.07057955529954699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,16384,0.2649875481923421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,65536,0.9794222513834635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,4096,0.05952178107367622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,12288,0.2054142157236735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,3584,0.029218667083316382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,4096,0.03244888782501221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,3584,0.05465066764089796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,3072,0.02573866645495097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,3072,0.050183110766940646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,10240,0.17214577727847627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,2560,0.022424888279702928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,2560,0.04243555665016174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,8192,0.13961866166856554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,2048,0.018769777483410306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,2048,0.037238223685158625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,7168,0.12414222293429905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,1536,0.015303111738628812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,1536,0.03187555405828688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,6144,0.10884533325831096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,1024,0.011728000309732227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,5120,0.09607644213570489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,1024,0.02475555572244856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,768,0.00998933364947637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,768,0.023065777288542852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,3584,0.07732977469762166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,512,0.007800000409285228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,512,0.02015555567211575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,2560,0.06215733289718628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,256,0.006137777947717243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,256,0.018178666631380718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,4096,0.08254400226804945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,128,0.00526311124364535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,128,0.017077333397335477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,768,0.040030221144358315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,64,0.0047173334492577445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,64,0.01737955543729994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,32,0.004975111120276981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,32,0.017117333081033494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,3072,0.07023821936713324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,1536,0.0499857763449351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,2048,0.05636355612013075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,16384,0.10099199745390151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,65536,0.3871768845452203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,65536,0.875641769833035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,16384,0.22667821248372397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,1024,0.0415013333161672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,12288,0.1723813348346286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,12288,0.07705866628222995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,10240,0.06502933634652032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,10240,0.1448728905783759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,8192,0.053391999668545194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,8192,0.10950844155417548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,128,0.03208977646297879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,7168,0.0473306675752004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,256,0.03344000048107571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,7168,0.09637155797746445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,6144,0.041491554843054876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,6144,0.08355110883712769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,512,0.03491466575198703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,5120,0.03522133496072557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,5120,0.07101244396633573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,65536,0.9664328893025717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,4096,0.029263112280103896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,4096,0.059176001283857554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,16384,0.2621777852376302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,3584,0.02661955522166358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,12288,0.20295822620391846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,3584,0.05413600140147739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,3072,0.023563555545277063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,3072,0.049582223097483315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,10240,0.17008088694678414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,2560,0.02040088839001126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,2560,0.041675554381476514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,8192,0.13815644052293566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,2048,0.017321777012613084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,2048,0.03639111253950331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,7168,0.12350222799513076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,1536,0.014474666780895658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,1536,0.03171822097566392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,6144,0.10768889056311713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,1024,0.011402666568756104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,1024,0.024285333024130926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,5120,0.09487555424372356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,768,0.009916444619496664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,768,0.02196888956758711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,3584,0.07624266544977824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,512,0.007885333564546373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,512,0.019722667005327012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,2560,0.06109155548943413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,256,0.0063777777055899305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,256,0.017753778232468497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,3072,0.06970489025115967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,128,0.005548444473081165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,128,0.016925333274735343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,4096,0.08239910999933879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,64,0.005153777698675792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,64,0.017277333471510146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,32,0.005437333136796951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,32,0.017044444878896076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,2048,0.05507911245028178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,1536,0.04981866810056898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,768,0.04054933455255296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,65536,0.3492284350925022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,65536,0.8716382450527616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,16384,0.08735199769337972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,16384,0.22506756252712676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,12288,0.06868355804019503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,12288,0.1725217766231961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,1024,0.041287110911475286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,10240,0.14431377251942953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,10240,0.06453688939412434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,8192,0.046485334634780884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,8192,0.10879288779364692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,128,0.03196977906756931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,256,0.03291733397377862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,7168,0.09442755248811509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,7168,0.04238577683766683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,6144,0.038797332180870905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,6144,0.08287644386291504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,512,0.035996443695492215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,5120,0.03333688775698344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,5120,0.0701191094186571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,16384,0.26469066407945424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,4096,0.02593333356910282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,4096,0.05938400162590874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,65536,0.9771644804212781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,3584,0.022844443718592327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,3584,0.05400088760587904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,12288,0.20224444071451822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,3072,0.02158311174975501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,3072,0.04931644598642985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,10240,0.16947021749284533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,2560,0.0185253338681327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,2560,0.041139556301964655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,8192,0.1378213299645318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,2048,0.015308444698651632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,2048,0.036151111125946045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,7168,0.12202133072747125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,1536,0.012454222473833295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,6144,0.10684000121222602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,1536,0.031081779135598078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,5120,0.09480088949203491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,1024,0.009759999811649323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,1024,0.024688889582951862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,768,0.008488000267081791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,768,0.02160177793767717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,3584,0.07597421937518649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,4096,0.08244888650046454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,512,0.006807111203670502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,512,0.019824888971116807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,256,0.005462222215202119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,256,0.0177848885456721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,3072,0.06934044758478801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,2560,0.060883555147382945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,128,0.004719999929269155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,128,0.016857778032620747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,64,0.004249777644872665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,64,0.01703644461101956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,32,0.004469333423508538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,32,0.016833777228991192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,2048,0.054951111475626625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,65536,0.7197235425313314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,65536,0.308676454755995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,16384,0.0776453349325392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,16384,0.18725421693589953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,1536,0.04942044284608629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,12288,0.05896977583567301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,12288,0.14504978391859266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,1024,0.0415635539425744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,768,0.03812088900142246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,10240,0.11936355961693658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,10240,0.05030666788419088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,8192,0.041626665327284075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,8192,0.09067110882865058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,256,0.03353777858946059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,128,0.032152000400755144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,512,0.03475822342766656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,7168,0.07926044199201795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,6144,0.03169688913557265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,7168,0.0372631119357215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,6144,0.07008977731068929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,5120,0.02735111117362976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,5120,0.05997244517008463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,16384,0.24553688367207846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,4096,0.023000889354281958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,4096,0.050160000721613564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,65536,0.8994133207533094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,3584,0.0207022229830424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,3584,0.046055999067094594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,12288,0.18919644090864393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,3072,0.017991110682487488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,3072,0.04193688763512505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,10240,0.15837866730160183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,2560,0.0158595558669832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,2560,0.035271111461851336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,8192,0.1288951105541653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,2048,0.013619555367363824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,2048,0.031325333648257785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,6144,0.10104444291856553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,1536,0.011360000405046674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,1536,0.028536889288160536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,5120,0.08878755569458008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,1024,0.008882666627566019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,7168,0.11464444796244304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,1024,0.022151110900772944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,768,0.007716444631417592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,768,0.020280889338917203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,4096,0.07789244254430135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,512,0.006211555666393704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,512,0.018411555223994784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,3584,0.07195822397867839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,256,0.005079999979999331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,256,0.01686755485004849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,3072,0.0660275552007887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,128,0.004416888786686791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,2560,0.057257778114742704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,128,0.01587199999226464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,64,0.00425866660144594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,64,0.0158933334880405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,32,0.004320000194840961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,32,0.015596444408098856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,2048,0.05153510967890421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,65536,0.7001662254333496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,65536,0.23370220926072863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,16384,0.0951075553894043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,1536,0.047304888566335045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,768,0.03713066710366143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,16384,0.18590488698747423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,12288,0.1414639949798584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,1024,0.0402782228257921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,12288,0.07179466883341472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,10240,0.11624444855584039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,10240,0.061287111706203885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,8192,0.047064890464146934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,8192,0.09050577878952026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,256,0.032425777779685125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,512,0.03306844499376085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,7168,0.043327109681235425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,7168,0.07821333408355713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,128,0.03128977616628011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,6144,0.03661244445376926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,6144,0.06872355275683932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,5120,0.031170666217803955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,5120,0.05881333351135254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,65536,0.9023528628879123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,4096,0.02087822225358751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,4096,0.04965600040223864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,16384,0.24734134144253203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,10240,0.15947643915812174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,3584,0.019352888067563374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,12288,0.19017244709862602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,3584,0.04569333460595873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,3072,0.01716977854569753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,3072,0.041526221566730075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,2560,0.01479022204875946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,2560,0.03584533267550998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,8192,0.12855911254882812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,2048,0.01237422227859497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,2048,0.031856000423431396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,7168,0.11429333686828613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,1536,0.010250666903124915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,6144,0.10074044598473443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,1536,0.027041777968406677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,1024,0.007937777373525832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,1024,0.02181333303451538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,5120,0.08937866820229425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,768,0.0064293332397937775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,768,0.01996977792845832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,3584,0.07223555776807997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,512,0.005431111074156231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,512,0.018301332990328472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,2560,0.05858844518661499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,256,0.004551110996140374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,256,0.016703999704784818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,4096,0.0779822203848097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,128,0.003995555556482739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,128,0.01572177807490031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,3072,0.06505333052741157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,64,0.003645333151022593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,2048,0.0517119997077518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,64,0.015655999382336933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,32,0.003899555653333664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,32,0.01555999947918786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,65536,0.7007155418395996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,65536,0.1811599996354845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,768,0.03637689020898607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,16384,0.18338221973843047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,16384,0.05679111348258125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,1536,0.047228445609410606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,12288,0.14133155345916748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,12288,0.038669334517584905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,10240,0.03305600086847941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,10240,0.11682132879892985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,1024,0.03950577643182542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,8192,0.027585777971479628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,8192,0.09018933110766941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,256,0.0321084459622701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,7168,0.024315555890401203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,128,0.030838221311569214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,7168,0.07800355884763929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,6144,0.021163556310865615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,6144,0.06789333290523954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,512,0.032913777563307024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,5120,0.018391110830836825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,5120,0.058415108256869845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,16384,0.24322843551635742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,4096,0.015504888362354703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,4096,0.04935377836227417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,65536,0.8833600150214301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,3584,0.01404622197151184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,3584,0.04505955510669284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,12288,0.18658222092522514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,3072,0.012647999657524956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,3072,0.04085066583421495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,10240,0.15638932916853163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,2560,0.011201777391963534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,2560,0.0354746679464976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,8192,0.1269528865814209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,2048,0.012655110822783576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,2048,0.031622221072514854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,7168,0.11378933323754205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,1536,0.009919111099508073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,1536,0.026750221848487854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,6144,0.09954666429095799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,1024,0.007598222129874759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,1024,0.02160800000031789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,5120,0.08830755286746556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,768,0.006527110934257507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,768,0.019851555426915485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,4096,0.07720266448126899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,512,0.005382222019963794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,3584,0.07077778047985502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,512,0.018135999639829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,256,0.004322666674852371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,256,0.016566221912701923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,3072,0.06428977515962389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,128,0.003744889050722122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,128,0.015579554769727917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,2560,0.05744889047410753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,2048,0.05110933383305868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,64,0.0036355555057525635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,64,0.015728889240158927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,32,0.0037813331517908308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,32,0.015596444408098856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,65536,0.6908355818854438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,65536,0.12565778361426458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,16384,0.04314666655328539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,16384,0.18488177988264295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,1536,0.047157333956824414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,12288,0.05283555719587538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,12288,0.14000266128116184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,1024,0.03916533456908332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,10240,0.04172622164090475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,10240,0.1159306632147895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,768,0.03851022322972616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,8192,0.035939554373423256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,8192,0.08912355369991726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,256,0.03168977631462945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,7168,0.03126488791571723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,512,0.03263377812173631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,7168,0.07881777816348605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,128,0.03050755461057027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,6144,0.02661688956949446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,6144,0.06780178017086454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,5120,0.019544000426928203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,5120,0.058387557665506996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,16384,0.23203023274739584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,65536,0.8446284400092231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,4096,0.014846222268210517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,4096,0.048792888720830284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,3584,0.012811555630630918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,3584,0.044440001249313354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,12288,0.178802662425571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,3072,0.01145066652033064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,3072,0.04038577940728929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,10240,0.14978755844963923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,2560,0.010379555324713389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,2560,0.0343848897351159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,8192,0.12143466207716201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,2048,0.008656000097592672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,2048,0.030751112434599135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,7168,0.10803288883633083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,1536,0.007117333511511485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,1536,0.025713778204388086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,6144,0.0959822204377916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,1024,0.00573511090543535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,1024,0.02146133283774058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,5120,0.08425511254204644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,768,0.005019555489222209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,768,0.0195795562532213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,3584,0.06734666559431288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,512,0.004410666724046071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,512,0.018210666047202218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,4096,0.07316977447933622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,256,0.0038497779104444715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,256,0.016365332735909354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,2560,0.055233776569366455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,128,0.0035457776652442086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,128,0.015439111325475903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,3072,0.06174755758709378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,64,0.0032640000184377036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,64,0.015368888775507608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,32,0.003391999958289994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,32,0.015240889456537036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,65536,0.10471377770105998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,2048,0.048467556635538735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,65536,0.6875626775953504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,16384,0.034308443466822304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,16384,0.1831973393758138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,768,0.03459022111362881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,12288,0.02736977736155192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,12288,0.1386853324042426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,1536,0.04364000095261467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,10240,0.02331999937693278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,10240,0.11499910884433323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,1024,0.03542755709754096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,8192,0.019254222512245178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,8192,0.08937066793441772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,256,0.031269331773122154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,7168,0.016965332958433364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,7168,0.07768444220225017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,512,0.031874666611353554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,6144,0.014986667368147107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,6144,0.06691110796398587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,128,0.030273778571022883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,5120,0.013596444494194455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,5120,0.05759288867314657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,16384,0.22494488292270234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,4096,0.011032889286677042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,4096,0.04868888854980469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,65536,0.8228178024291992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,3584,0.01036622209681405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,3584,0.04420800010363261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,12288,0.17330043845706514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,3072,0.009222221871217092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,3072,0.039461334546407066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,10240,0.14630311065249973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,2560,0.011528888510333167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,2560,0.034400890270868935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,8192,0.11843466758728027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,2048,0.009791110952695211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,2048,0.029614223374260798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,7168,0.10543733172946507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,1536,0.008147555920812819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,1536,0.026514665948020086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,6144,0.09343466493818496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,1024,0.006257777826653586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,1024,0.021158221695158217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,5120,0.08302755488289727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,768,0.005513777749405966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,768,0.019380443625979953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,3584,0.06664799981647067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,512,0.004687110996908612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,4096,0.07182844479878743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,512,0.017699556218253244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,256,0.003863111138343811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,256,0.01610311037964291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,3072,0.05989155504438612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,128,0.003551111039188173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,128,0.01521511044767168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,2560,0.05379911263783773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,64,0.0033297776761982176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,64,0.015246222416559855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,32,0.0032951111594835916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,32,0.015078221758206686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,65536,0.0781608886188931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,65536,0.6888586680094401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,2048,0.047853334082497485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,16384,0.027235555979940627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,16384,0.18182844585842559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,1536,0.04452177882194519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,12288,0.021239999267790053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,12288,0.13753422101338705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,768,0.03520800007714166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,1024,0.03597777750757005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,10240,0.020943111843532983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,10240,0.11586933665805393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,8192,0.016336888074874878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,8192,0.08923555745018853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,256,0.030862222115198772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,128,0.03014666504330105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,512,0.03147555390993754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,7168,0.015024888846609326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,7168,0.07765689161088732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,6144,0.013398222625255585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,6144,0.06713244650099012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,5120,0.011885333392355176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,5120,0.057767113049825035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,65536,0.8109262254503039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,4096,0.01016088906261656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,4096,0.04832177691989475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,16384,0.22155022621154785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,3584,0.009151111046473185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,3584,0.04375911090109083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,12288,0.1718604432211982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,3072,0.008407110969225565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,3072,0.039416889349619545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,10240,0.1432853274875217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,2560,0.007771555748250749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,2560,0.03360711203681098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,8192,0.11570221847958034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,2048,0.0063733334342638654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,2048,0.030382222599453394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,7168,0.10343199968338013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,6144,0.09219200081295437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,1536,0.005594666633341048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,1536,0.025079111258188885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,1024,0.0047884442740016514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,5120,0.081185777982076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,1024,0.020985777179400127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,768,0.004344888859324985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,768,0.019337778290112812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,4096,0.07019466824001737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,512,0.003949333396222857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,512,0.017567111386193168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,3584,0.06456355253855388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,256,0.00371111101574368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,256,0.015860444969601102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,3072,0.059239996804131397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,128,0.003343111111058129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,128,0.015258666541841296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,2048,0.045984887414508395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,2560,0.05269333389070299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,64,0.003157333367400699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,32,0.003314666656984223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,64,0.015304888288180033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,32,0.015035554766654968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,65536,0.058596445454491504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,65536,0.685716470082601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,16384,0.01776000029510922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,16384,0.18350311120351157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,1536,0.04342755675315857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,12288,0.01626311077011956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,12288,0.13754399617513022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,768,0.03433955709139506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,1024,0.0350524451997545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,10240,0.014797333214018079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,10240,0.11495021979014079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,8192,0.012822222378518848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,8192,0.08921244409349229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,512,0.031217777066760596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,7168,0.01202666676706738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,7168,0.07751288678910997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,256,0.03088266650835673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,6144,0.012138666378127204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,6144,0.06682844294442071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,128,0.030215111043718126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,5120,0.01054044481780794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,5120,0.05715555614895291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,16384,0.21436799897087944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,65536,0.7804462114969889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,4096,0.008715555899673039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,4096,0.04807910985416836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,3584,0.008049777812427944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,3584,0.04355288876427544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,10240,0.13889155122968885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,12288,0.16540443897247314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,3072,0.007335111498832703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,3072,0.03826311230659485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,2560,0.0069982219073507525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,2560,0.03287377622392442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,8192,0.11221333344777425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,2048,0.00600355573826366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,2048,0.028304888142479792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,7168,0.10029777553346421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,1536,0.005519111123349931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,1536,0.02518133322397868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,6144,0.08895022339291042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,1024,0.004540444248252445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,1024,0.020987555384635925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,5120,0.07933421929677327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,768,0.004322666674852371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,768,0.01960177719593048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,4096,0.06782578097449408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,512,0.0037768888804647657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,512,0.017619555195172627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,3584,0.06344533628887601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,256,0.003474666840500302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,256,0.015956444872750174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,3072,0.057242665025922984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,128,0.0031759999692440033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,128,0.01519555515713162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,2560,0.05096711052788628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,64,0.0030462222380770575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,64,0.015079110860824585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,32,0.003080000066094928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,32,0.014879999889267815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,65536,0.05652533637152778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,65536,0.6854471100701226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,2048,0.043699555926852755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,16384,0.011710222396585675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,16384,0.18314488728841147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,1536,0.04003555576006571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,12288,0.0123511114054256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,12288,0.1382675568262736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,1024,0.03411555621359084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,10240,0.011023999916182624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,10240,0.11601955360836452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,768,0.03377422359254625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,8192,0.009896888501114314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,8192,0.08959466881222194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,512,0.030903998348448012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,7168,0.009401777552233802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,256,0.030743112166722614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,7168,0.07744710975223117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,6144,0.008811555802822113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,6144,0.06660977999369304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,128,0.029772443903817072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,5120,0.008752000000741746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,5120,0.05712266763051351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,65536,0.7773422135247124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,4096,0.007999999655617608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,4096,0.047269334395726524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,16384,0.2131333351135254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,3584,0.007792889244026608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,3584,0.04296000136269463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,12288,0.1651715570025974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,3072,0.007094221810499827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,3072,0.03829333186149597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,10240,0.13863821824391684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,2560,0.0063777777055899305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,2560,0.03259733319282532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,8192,0.11254844400617812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,2048,0.005753777921199799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,2048,0.02836266822285122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,7168,0.1003120011753506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,1536,0.005028444445795483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,1536,0.02500711050298479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,6144,0.08834399779637654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,1024,0.004305777864323722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,1024,0.02097333305411869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,5120,0.07819377713733248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,768,0.003949333396222857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,768,0.019140443868107267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,4096,0.06812000274658203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,512,0.0036275556517971885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,512,0.017196445001496207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,3584,0.062237335575951465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,256,0.0032702222880389956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,3072,0.05683910846710205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,256,0.01564800077014499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,2560,0.05096355411741468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,128,0.003042666655447748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,128,0.014954666296641031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,64,0.002893333426780171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,64,0.014897776974572076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,32,0.0029120000286234748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,32,0.014683556225564746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,65536,0.047425776720047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,65536,0.686869356367323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,16384,0.013008000122176277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,16384,0.18332710531022814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,12288,0.010800888968838586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,12288,0.13956444793277317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,10240,0.009889778163697984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,10240,0.11597777737511529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,8192,0.008781332936551835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,8192,0.08925866418414646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,7168,0.008417777717113495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,7168,0.07730222410625882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,6144,0.007859555383523306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,6144,0.06683111190795898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,5120,0.007435555259386699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,5120,0.056806219948662647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,4096,0.007039111521508958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,4096,0.04723733332422045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,3584,0.007376000285148621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,3584,0.0429057776927948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,3072,0.007121777368916406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,3072,0.03731377919514974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,2560,0.0063235556913746735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,2560,0.031665778822369046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,2048,0.005596444424655702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,2048,0.02796799937884013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,1536,0.004914666629499859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,1536,0.024801777468787298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,1024,0.00423200014564726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,2048,0.043175998661253184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,768,0.003986666599909465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,1024,0.02071822186311086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,512,0.0035946667194366455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,768,0.01928444372283088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,256,0.0032879999942249725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,512,0.017215111189418368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,128,0.002987555538614591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,256,0.015816888875431485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,64,0.002943111169669363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,128,0.014840889308187695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,32,0.002858666703104973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,64,0.01499288943078783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,32,0.014675555957688225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,65536,0.046006222565968834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,16384,0.010040000081062317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,16384,0.18382844660017225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,65536,0.6869422064887153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,12288,0.008955555657545725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,1536,0.04009244508213467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,10240,0.008278222547637092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,8192,0.007698666718271043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,12288,0.13944889439476862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,10240,0.11594577630360921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,1024,0.03447022371821933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,7168,0.007048889166778988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,8192,0.08962044450971816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,6144,0.0069022224181228215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,7168,0.0771920018725925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,5120,0.007328888608349695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,6144,0.06652888986799452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,4096,0.006928000185224745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,5120,0.05638844437069363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,3584,0.007270221908887227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,4096,0.04781244529618157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,3072,0.0070426662762959795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,3584,0.04310755597220527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,2560,0.00628977765639623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,3072,0.03706399930848016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,2048,0.005552000055710475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,2560,0.032856888241238065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,1536,0.004874666531880696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,2048,0.02828622195455763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,1024,0.0041982221106688184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,1536,0.02498666610982683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,768,0.0038844446341196695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,1024,0.02077511118517982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,512,0.0035546666218174826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,768,0.01922044489118788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,256,0.0031635556370019913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,768,0.033415112230512835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,256,0.01591466698381636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,512,0.01759111053413815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,128,0.002971555623743269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,128,0.014894222219785055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,64,0.0028320000403457214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,32,0.002899555489420891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,64,0.01516888870133294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,32,0.014879110786649915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,512,0.030589332183202107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,256,0.03050666716363695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,128,0.029847111966874864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,12288,0.6798897849188911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,12288,0.7000373204549154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,16384,0.8857350879245334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,10240,0.5757199923197428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,16384,0.8963297737969292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,10240,0.7673938009474012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,12288,1.0125012927585177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,8192,0.4382302231258816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,10240,0.5740462409125434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,16384,1.3115840488009982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,8192,0.46326578987969297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,8192,0.6036613252427843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,7168,0.3882764445410834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,6144,0.3431768947177463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,5120,0.2773004372914632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,6144,0.37360265519883895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,7168,0.42572887738545734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,5120,0.30552445517645943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,5120,0.39119201236301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,7168,0.5588480101691352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,4096,0.22625244988335502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,3584,0.19528710842132568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,4096,0.25291376643710667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,6144,0.46353779898749453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,3072,0.16522488329145643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,3584,0.23500355084737143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,3072,0.20117333200242785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,4096,0.3167191081576877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,3584,0.28254932827419704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,2560,0.1524160040749444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,2560,0.21192355950673422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,2560,0.17788355880313447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,2048,0.11774578359391953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,3072,0.2470782332950168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,1536,0.09723289145363702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,1024,0.07713244358698527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,2048,0.15249778164757624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,2048,0.1709137757619222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,1536,0.122688889503479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,1024,0.09699910879135132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,768,0.06686222553253174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,512,0.05816266934076945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,512,0.07379021909501818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,256,0.04804177747832405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,1024,0.10241599877675374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,512,0.06097155809402466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,256,0.04390755626890394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,128,0.036814222733179726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,1536,0.13225066661834717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,256,0.06495733393563165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,768,0.08578488561842178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,128,0.04387199878692627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,64,0.03185333477126227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,32,0.032418668270111084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,64,0.04617244336340162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,128,0.06260710954666138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,32,0.05123466584417555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,768,0.07981689108742608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,12288,0.2126177814271715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,16384,0.28062934345669216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,16384,0.32159556282891166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,12288,0.22647288110521105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,10240,0.18127200338575575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,12288,0.2480177879333496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,10240,0.19148711363474527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,16384,0.29728711975945366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,10240,0.22119200229644775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,8192,0.1396177742216322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,7168,0.12451555993821885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,8192,0.15539022286732992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,8192,0.17528977659013537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,6144,0.11083110835817124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,7168,0.13640711042616102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,7168,0.15848178333706325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,6144,0.1403813362121582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,6144,0.11873688962724473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,5120,0.09422133366266887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,5120,0.12314044104682074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,4096,0.08213688929875691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,4096,0.0760257773929172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,5120,0.10068621900346543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,3584,0.06939822435379028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,4096,0.10391110844082302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,3584,0.09603733486599392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,3072,0.06462755468156603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,3072,0.06229777468575371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,3584,0.07430577940411039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,3072,0.08645777569876777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,2560,0.055919110774993896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,2560,0.05580711364746094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,2560,0.08030577500661214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,2048,0.04558222161398994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,1536,0.03889866669972738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,2048,0.068086220158471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,2048,0.04624977707862854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,1536,0.059106667836507164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,1024,0.03166044420666165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,1536,0.03681777914365133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,768,0.022427555587556627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,1024,0.050198223855760365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,1024,0.027297778262032404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,512,0.01658933361371358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,768,0.028184887435701158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,512,0.025797333982255723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,256,0.01202488856183158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,768,0.04633511106173197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,65536,1.1294551425509982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,512,0.043403555949529014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,256,0.02235377828280131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,128,0.009458666874302758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,256,0.03883466786808438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,65536,1.252804438273112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,64,0.008033778104517195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,128,0.020012444920010038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,32,0.00814133303032981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,64,0.020297777321603563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,128,0.038719111018710665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,32,0.022357333037588332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,16384,0.2332977718777127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,16384,0.2778906557295057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,16384,0.22758844163682726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,12288,0.17769155237409803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,12288,0.17490932676527235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,12288,0.21217421690622965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,10240,0.14890666802724203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,10240,0.14580621984269884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,10240,0.18099378214942086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,8192,0.11853244569566514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,8192,0.11725155512491862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,8192,0.150438215997484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,7168,0.10358311070336236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,7168,0.136790222591824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,7168,0.10605421993467544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,6144,0.0930666658613417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,6144,0.09109777874416775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,5120,0.0800328916973538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,6144,0.12239289283752441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,65536,1.0656382242838542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,65536,0.934817738003201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,5120,0.10621511273913914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,5120,0.07825422286987305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,4096,0.0649155577023824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,4096,0.0961573322614034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,3584,0.058280891842312284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,4096,0.06379110945595635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,3584,0.05707377857632107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,3072,0.05228088961707222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,3584,0.08307466904322307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,3072,0.0779048866695828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,2560,0.04890222350756327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,2560,0.043106665213902794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,3072,0.05024888780381945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,2560,0.070433775583903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,65536,1.2129857805040147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,2048,0.03928088810708787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,1536,0.033486220571729876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,2048,0.05974488788180881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,1024,0.024549333585633173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,1024,0.028166221247778997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,1536,0.05284888876809014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,768,0.016907556189431083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,512,0.013390222357379066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,1024,0.04648088746600681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,1536,0.028412444723976985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,512,0.023208000593715247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,2048,0.03602755400869582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,768,0.042845331960254245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,256,0.009576000273227692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,768,0.025416889124446448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,256,0.02030311193731096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,128,0.018366222580273945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,128,0.007679111427730984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,512,0.04012177719010247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,256,0.037294222248925105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,64,0.006544888847404056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,32,0.006807111203670502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,128,0.03420799970626831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,64,0.01881511178281572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,32,0.019428445233239066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,16384,0.22485065460205078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,65536,1.0209884643554688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,12288,0.14864533477359348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,16384,0.1910897758271959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,12288,0.16887377368079293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,10240,0.14148355854882136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,10240,0.12449333402845596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,8192,0.10024711158540513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,8192,0.11299288272857666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,65536,0.9061039818657769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,7168,0.09018488725026448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,7168,0.10499733024173313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,6144,0.07886488570107354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,6144,0.08973244163725112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,65536,0.7890488836500379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,5120,0.0666524436738756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,5120,0.08163111077414618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,16384,0.25908621152242023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,4096,0.054851555162005954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,4096,0.06365333663092719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,12288,0.20002133316463896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,10240,0.17143732971615264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,3584,0.049689776367611356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,3584,0.05760888920889961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,3072,0.04418666826354133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,3072,0.05127999848789639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,8192,0.14215289221869573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,2560,0.037997331884172224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,2560,0.045062220758861966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,65536,0.9523617426554362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,7168,0.1305333375930786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,2048,0.03772622346878052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,2048,0.03250133328967624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,1536,0.028993778758578833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,1536,0.03291022115283542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,6144,0.11650578180948894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,1024,0.01849333279662662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,1024,0.027992000182469685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,5120,0.10223555564880371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,768,0.015084443820847405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,768,0.02515911062558492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,4096,0.08644177516301473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,512,0.011911110745535957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,512,0.022856889499558344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,3584,0.07972888814078437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,256,0.008847999903890822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,256,0.01958311100800832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,2560,0.06683111190795898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,128,0.007035555938879649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,128,0.018379555808173288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,3072,0.07268355290095012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,64,0.006169777777459886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,64,0.01885066595342424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,32,0.0064311110311084324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,32,0.019070222973823547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,768,0.04214844438764784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,1536,0.051623112625545926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,2048,0.05812888675265842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,1024,0.045010666052500405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,65536,0.7348666720920138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,16384,0.15209689405229357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,16384,0.18618488311767578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,16384,0.23385598924424914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,65536,0.8489191267225477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,12288,0.1162124474843343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,12288,0.14084711339738634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,12288,0.17938133080800375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,65536,0.6143831147087945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,10240,0.11804621749454075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,10240,0.15253778298695883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,8192,0.09905688630210029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,10240,0.09939199686050415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,8192,0.07986755503548516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,128,0.03382488754060533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,7168,0.0846524437268575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,8192,0.12704532676272923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,256,0.03645155496067471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,7168,0.11678044001261394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,6144,0.07523733377456665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,6144,0.10392177767223781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,6144,0.06133777565426297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,7168,0.07095999850167169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,5120,0.06579466660817464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,3584,0.04862933357556661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,5120,0.052689777480231396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,4096,0.053929779264662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,5120,0.09208444754282634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,4096,0.042945779032177396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,4096,0.07923911015192668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,3584,0.07261066966586642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,3072,0.04388533367051018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,2560,0.04177600145339966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,3072,0.06707200076844957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,2560,0.06309155623118083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,2048,0.032157331705093384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,3584,0.039149334033330284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,1536,0.01978222197956509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,2048,0.053245332505967885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,2048,0.025098666548728943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,2560,0.029976000388463337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,1536,0.02819733487235175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,1024,0.01461866663561927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,3072,0.03465244505140517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,1024,0.024420445164044697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,1024,0.04157422317398919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,1536,0.04651555418968201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,768,0.012449777788586087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,768,0.022329777479171753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,512,0.01019022199842665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,768,0.040038221412234835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,512,0.02086755633354187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,256,0.007484444313579136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,512,0.03707733419206407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,256,0.017902221944597032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,128,0.006072888771692912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,256,0.032813333802753024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,128,0.017088888419999015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,64,0.005450666778617435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,128,0.0326551099618276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,32,0.0058355554938316345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,64,0.017179555363125272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,32,0.0173653331067827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,512,0.038647111919191145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,16384,0.18065244621700713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,16384,0.23208623462253145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,12288,0.1350133286582099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,12288,0.11408355500962998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,16384,0.15160977840423584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,10240,0.11198400126563178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,12288,0.17698578039805093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,10240,0.09666311078601414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,10240,0.15021333429548475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,8192,0.0914728906419542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,8192,0.0783768892288208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,8192,0.12583466370900473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,7168,0.08296089039908515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,7168,0.11332089371151394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,7168,0.06991822189754911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,6144,0.07354755534066094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,6144,0.10250577661726211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,5120,0.06440533532036676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,4096,0.05236355463663737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,5120,0.058703998724619545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,6144,0.06346133020189074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,4096,0.07739110787709554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,3584,0.047635555267333984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,5120,0.08998933103349473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,4096,0.03971644573741489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,65536,0.7006666925218371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,3072,0.042745778958002724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,3584,0.07148177756203546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,2560,0.04064355625046624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,3072,0.06551555792490642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,3584,0.034853332572513156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,2560,0.06172355678346422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,2048,0.021176889538764954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,2048,0.032300445768568255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,1536,0.01772799922360314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,3072,0.03197333216667175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,2048,0.05286488930384318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,1536,0.028063999281989202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,1024,0.012888888518015543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,1536,0.046874665551715426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,2560,0.02720888952414195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,1024,0.02424533334043291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,768,0.010457778142558204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,1024,0.04235555397139656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,768,0.02201777696609497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,512,0.00830577810605367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,768,0.04015200005637275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,512,0.020644444558355544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,256,0.006244444598754247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,512,0.03791466686460707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,256,0.017886221408843994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,128,0.005295111073387994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,128,0.01697422232892778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,256,0.03346666693687439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,64,0.004979555391603046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,32,0.005256888767083486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,128,0.03268977668550279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,64,0.017133333616786532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,32,0.017313778400421143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,65536,0.8507591353522407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,16384,0.17851644092135957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,16384,0.20652800136142305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,16384,0.115264892578125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,12288,0.13246933619181314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,65536,0.5877324210272895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,12288,0.08822133143742879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,10240,0.11095022492938572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,12288,0.15840089321136475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,10240,0.142812450726827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,10240,0.07417600022421943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,8192,0.09162488910886978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,8192,0.06333421998553805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,8192,0.11417510774400498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,7168,0.08209066920810275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,7168,0.05449600021044413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,7168,0.10460178057352702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,6144,0.07253511084450616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,5120,0.0633902218606737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,6144,0.09421600235833062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,5120,0.08356000317467584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,6144,0.04723111126157972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,4096,0.05173777871661716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,5120,0.040115555127461754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,4096,0.07218488719728258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,3584,0.04672266708479988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,4096,0.033504890071021184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,3072,0.0440942214594947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,3584,0.0659466650750902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,65536,0.7069839901394315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,3072,0.062296887238820396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,2560,0.02314666741424137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,3584,0.030239999294281006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,2560,0.03862133291032579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,2048,0.01906044450071123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,65536,0.7535671128167046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,2048,0.0321066677570343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,3072,0.026686223016844854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,2560,0.05801777707205879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,1536,0.015474667151769003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,1536,0.028246220615175035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,1024,0.011952888634469775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,2048,0.04907911022504171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,768,0.010271111296282874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,1024,0.024145777026812237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,1536,0.04324000080426534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,1024,0.03922755519549052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,768,0.02212000058756934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,512,0.008420444197124904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,768,0.03778310947948032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,512,0.019722667005327012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,256,0.0063031112982167145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,512,0.035164445638656616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,256,0.017972444494565327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,128,0.005220444252093633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,256,0.03179733289612664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,128,0.016975111431545682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,64,0.004639111045334074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,32,0.004955555415815777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,128,0.031160887744691636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,64,0.017089777522616915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,32,0.017072889539930556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,16384,0.17768000231848824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,16384,0.19978933864169648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,16384,0.1004062228732639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,12288,0.13353689511617026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,12288,0.15453955862257215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,12288,0.07742222150166829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,10240,0.06556089056862725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,10240,0.1103093359205458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,10240,0.13233689467112222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,8192,0.05363644493950737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,8192,0.08976978063583374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,65536,0.443703121609158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,65536,0.7010817527770996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,8192,0.11112888654073079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,7168,0.08142222298516168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,7168,0.04811199837260776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,6144,0.07150399684906006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,5120,0.06269066863589816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,6144,0.09200000100665623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,7168,0.10186133119795059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,5120,0.08136622111002605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,4096,0.050981332858403526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,6144,0.0421706669860416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,4096,0.06999822457631429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,5120,0.03602577911482917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,3584,0.04602933261129591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,3072,0.024263999528355066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,4096,0.03007022208637661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,3072,0.04290666513972812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,3584,0.06495910882949829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,3072,0.06102755334642199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,2560,0.020937777227825586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,3584,0.027360000544124182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,2560,0.037406222687827215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,2048,0.01780533293883006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,2048,0.031558222240871854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,2560,0.05664622121387058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,1536,0.014380445082982382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,1536,0.027656000521447923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,1024,0.011350221931934357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,2048,0.04816177818510267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,1024,0.02360533343421088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,1536,0.04437511165936788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,768,0.021651556094487507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,1024,0.03821511069933573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,768,0.009911999934249455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,512,0.008212444682916006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,256,0.006302222195598815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,768,0.03704444567362467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,512,0.019323555959595572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,256,0.01744622157679664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,512,0.03341066506173875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,128,0.005079999979999331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,256,0.031361778577168785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,128,0.03181333343187968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,128,0.0165928883685006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,64,0.004586666822433472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,32,0.005060444275538127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,64,0.016935999194780987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,65536,0.7175635761684842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,32,0.016869333055284288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,16384,0.14167022705078125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,12288,0.06004799736870659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,16384,0.07875288857354058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,12288,0.10522577497694229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,10240,0.05137333273887634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,10240,0.08767466412650214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,65536,0.39203466309441465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,65536,0.5471208890279134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,8192,0.07240800062815349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,8192,0.04205155703756544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,7168,0.06648266977734037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,7168,0.03767022159364488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,6144,0.03321689036157396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,6144,0.05904710955089993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,65536,0.3077937761942546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,5120,0.028508444627126057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,5120,0.05199199914932251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,16384,0.188810666402181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,4096,0.023788443870014612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,4096,0.0422435568438636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,12288,0.1434613333808051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,3584,0.021024000313546922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,3584,0.03959733247756958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,10240,0.12395200464460586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,3072,0.01847555571132236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,3072,0.03545511100027297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,8192,0.10389688942167495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,2560,0.01607822212908003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,65536,0.6799866888258191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,2560,0.03189333279927572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,2048,0.013574221895800697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,7168,0.09456444448894924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,2048,0.027308444182078045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,1536,0.011361777782440186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,1536,0.02461066676510705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,6144,0.08510844575034247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,1024,0.00907199995385276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,1024,0.0214008887608846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,5120,0.07665510972340901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,768,0.007974222302436829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,768,0.01923288901646932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,4096,0.0653182201915317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,512,0.0061919999619325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,3584,0.06179910898208618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,512,0.017807111144065857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,256,0.005066666752099991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,256,0.016540444559521146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,3072,0.05626844697528415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,128,0.00444533344772127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,2560,0.05217511124081082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,128,0.01566844516330295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,64,0.004011555678314633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,64,0.015790222419632804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,32,0.004320000194840961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,32,0.015634665886561077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,2048,0.045801778634389244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,1536,0.0410791105694241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,1024,0.03740977909829881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,65536,0.5240355597601997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,16384,0.06889510816997953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,16384,0.1395368840959337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,65536,0.2748844358656141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,12288,0.1026115549935235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,12288,0.0537048876285553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,768,0.034852445125579834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,10240,0.08558044168684219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,10240,0.05148622062471178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,8192,0.037096887826919556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,8192,0.07107644610934787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,256,0.030975110001034204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,128,0.030415998564826116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,7168,0.0650906695259942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,512,0.03242044316397773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,7168,0.03272355596224467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,6144,0.05735466877619425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,6144,0.034330666065216064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,5120,0.027419555518362258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,5120,0.050275554259618126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,65536,0.7009938028123645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,4096,0.02055022286044227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,4096,0.04096800088882446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,16384,0.19026845031314424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,3584,0.018377777602937486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,3584,0.03866577810711331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,12288,0.14375466770595974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,3072,0.01645777788427141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,3072,0.03472888800832961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,10240,0.12275911702050103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,2560,0.014568888478808932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,2560,0.03178222311867608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,8192,0.10191822052001953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,2048,0.012315555579132505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,2048,0.027281777726279363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,7168,0.09322666459613377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,1536,0.010404444403118556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,1536,0.024538666009902954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,6144,0.08420444197124904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,1024,0.008344888687133789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,1024,0.020377778344684176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,5120,0.07499111360973783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,768,0.007017778025733099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,768,0.01866755551762051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,4096,0.06528444422615899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,512,0.005711110929648082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,512,0.01763288842307197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,3584,0.06125422318776449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,256,0.004826666580306159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,256,0.016506666938463848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,3072,0.05626933442221748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,128,0.004238222208287981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,128,0.015653333730167814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,2560,0.05256444546911451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,64,0.0038613333470291565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,64,0.01590666671593984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,32,0.004063111212518481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,32,0.015697777271270752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,2048,0.04612088865704007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,1536,0.041758222712410815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,768,0.034852445125579834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,1024,0.03793955511516995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,65536,0.5149217711554633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,65536,0.2388062212202284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,16384,0.13878844843970403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,16384,0.0649964412053426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,12288,0.10177689128451878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,12288,0.04900711112552219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,10240,0.08473422129948933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,10240,0.053605334626303784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,8192,0.033313777711656355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,8192,0.06993689139684041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,512,0.032808889945348106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,7168,0.03011911114056905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,7168,0.06346311171849568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,256,0.03089955449104309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,128,0.030954665607876245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,6144,0.05580088827345106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,6144,0.026026666164398193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,5120,0.022263111339675054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,5120,0.049808889627456665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,16384,0.18608889314863417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,12288,0.14155377282036677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,4096,0.018568888306617737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,4096,0.04059200154410468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,3584,0.01647111111217075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,65536,0.6705759896172417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,3584,0.037843555212020874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,3072,0.014631110760900708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,3072,0.034076445632510714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,10240,0.12166399425930446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,2560,0.012842666771676807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,2560,0.03139644530084398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,8192,0.10216266579098171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,2048,0.01108977778090371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,2048,0.027279110418425664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,7168,0.09370044204923843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,1536,0.009400889277458191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,1536,0.024217777782016333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,6144,0.0849368903372023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,1024,0.007426666716734569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,1024,0.02016266683737437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,5120,0.07547199726104736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,768,0.00638755535085996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,768,0.018754666050275166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,3584,0.060396446122063525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,512,0.005371555685997009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,512,0.017508443858888414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,2560,0.05163288778728909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,256,0.004559999952713649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,256,0.01633066601223416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,768,0.0341360006067488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,128,0.004028444488843282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,128,0.015620445211728414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,1536,0.041123555766211614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,64,0.003654222107595868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,64,0.015384889311260648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,32,0.0038648889296584656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,4096,0.06511021984948052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,32,0.015444444285498725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,3072,0.05628889136844211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,2048,0.045280890332327954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,65536,0.5226257642110189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,16384,0.13803733719719782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,16384,0.0625928905275133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,1024,0.036474668317370944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,12288,0.10149688853157891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,12288,0.054782221714655556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,65536,0.20788089434305826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,10240,0.08220177888870239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,10240,0.048471109734641186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,8192,0.034360888931486346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,8192,0.0692302253511217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,128,0.030335999197430078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,256,0.03038755390379164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,7168,0.03262844350602891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,7168,0.0620408919122484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,6144,0.02900622288386027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,6144,0.05573955509397718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,512,0.03207644489076402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,5120,0.025433777107132807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,5120,0.04910133282343546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,65536,0.6508720186021593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,16384,0.18301867114173043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,4096,0.01663466625743442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,4096,0.04007555709944831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,3584,0.016183111402723525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,3584,0.037090665764278836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,12288,0.1396826638115777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,3072,0.015722667177518208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,3072,0.0341635545094808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,10240,0.1197928852505154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,2560,0.013470222552617392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,8192,0.10029688808653091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,2560,0.03105866577890184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,2048,0.011224889092975192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,2048,0.02680177821053399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,7168,0.09226400322384304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,6144,0.08329777585135566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,1536,0.008969777988062965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,1536,0.023919110496838886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,1024,0.006769777586062749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,1024,0.02000177734427982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,5120,0.07457244396209717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,768,0.006234666539563074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,768,0.018552000323931377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,3584,0.059931556383768715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,512,0.005025777965784073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,512,0.017258667283587985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,2560,0.05150489012400309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,256,0.004368000146415499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,256,0.016470222009552848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,3072,0.055098666085137256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,128,0.003891555385457145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,128,0.015526221858130561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,4096,0.0642133355140686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,64,0.003549333247873518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,64,0.015602666470739575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,32,0.0036675557494163513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,32,0.01550044450494978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,2048,0.04500266578462389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,65536,0.16412177350785997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,65536,0.5119173261854384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,768,0.03319822086228265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,16384,0.050810668203565806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,16384,0.13797067271338567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,12288,0.050684445434146456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,12288,0.10078399711185032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,1536,0.0406951109568278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,10240,0.043014221721225314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,10240,0.08331288894017537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,1024,0.03664266732003953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,8192,0.03177777926127116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,8192,0.06812889046139188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,128,0.030329777134789362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,7168,0.028871999846564397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,7168,0.0621395574675666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,256,0.03007733490731981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,6144,0.02708622150950962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,6144,0.05433511071734958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,512,0.03161155515246921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,5120,0.02234399980968899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,5120,0.04846044381459554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,65536,0.6123458014594184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,4096,0.018031110366185505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,4096,0.04016000032424927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,16384,0.17240356074439156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,3584,0.017122666041056316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,12288,0.12823200225830078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,3584,0.036785778072145246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,3072,0.01556711064444648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,3072,0.03391999999682108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,10240,0.11011999845504761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,2560,0.012587555580668978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,2560,0.03110399842262268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,8192,0.09220000108083089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,2048,0.010750222537252637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,2048,0.026949332820044622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,7168,0.08512177732255723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,1536,0.008948444492287105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,1536,0.022865777214368183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,6144,0.07749955521689521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,1024,0.007041778001520369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,1024,0.020035554965337116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,5120,0.06926400131649442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,768,0.005688000056478713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,768,0.018420444594489205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,4096,0.0603440006573995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,512,0.005035555611054103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,512,0.017294221454196505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,3584,0.05582222011354235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,256,0.004208000169859992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,256,0.016233777006467182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,3072,0.05156888895564609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,128,0.0037439999481042228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,128,0.015490666031837463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,2560,0.04882755544450548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,64,0.0034177777253919174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,64,0.015398222539159985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,32,0.0035182221068276297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,32,0.01531733406914605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,65536,0.12643378310733372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,2048,0.043493333790037364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,65536,0.513603581322564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,16384,0.040707555082109236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,1536,0.03932799895604452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,16384,0.13636355929904515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,12288,0.04276355438762241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,12288,0.09827022420035468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,1024,0.03364355696572198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,10240,0.03724799884690179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,10240,0.0812693304485745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,768,0.03238577644030253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,8192,0.027746667464574177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,8192,0.06698044141133626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,512,0.031215111414591473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,7168,0.026037333740128413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,7168,0.059248891141679555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,256,0.029620445436901514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,128,0.030279109875361126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,6144,0.022822222775883142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,6144,0.05344177616967095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,5120,0.02036977807680766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,5120,0.04794933398564657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,16384,0.17355822192298043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,4096,0.015781333049138386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,4096,0.03970488905906677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,12288,0.13130666149987116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,3584,0.014188443620999655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,65536,0.6201848983764648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,3584,0.036525332265430026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,3072,0.012867555850081973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,3072,0.03339644604259067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,10240,0.11260888311598037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,2560,0.010823110739390055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,2560,0.030918220678965252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,8192,0.0942666663063897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,2048,0.009213333328564962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,2048,0.026268444127506677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,7168,0.08661688698662652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,1536,0.00792622235086229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,1536,0.022505778405401442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,6144,0.07831288708580865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,1024,0.0058986664646201665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,1024,0.019540444016456604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,5120,0.06994666655858357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,768,0.005158222383923001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,768,0.018408000469207764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,4096,0.060180445512135826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,512,0.004334222111437055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,512,0.017261332935757108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,3584,0.055083556307686694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,256,0.0038053331275780997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,256,0.016164445214801364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,3072,0.050810668203565806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,128,0.0034515555534097883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,128,0.015412444869677225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,2048,0.042740444342295326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,2560,0.048678222629759044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,64,0.0031555555760860443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,32,0.003301333429084884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,64,0.015438222222858004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,1536,0.0384880006313324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,32,0.015091554986106025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,768,0.032009777095582746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,65536,0.5138266881306967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,16384,0.13644888665941027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,16384,0.0315324432320065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,12288,0.03437422381507026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,12288,0.09808711210886638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,65536,0.09050489134258694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,10240,0.028833776712417603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,1024,0.03514044483502706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,10240,0.08159199688169691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,8192,0.020743111769358318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,8192,0.0670657753944397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,128,0.02973244587580363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,7168,0.017512889371977914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,7168,0.06054221921496921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,256,0.030188444587919448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,6144,0.01755644381046295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,6144,0.05323822299639384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,512,0.03122399912940131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,5120,0.016036444240146212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,5120,0.04730400111940172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,65536,0.5738586849636501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,4096,0.011857777833938599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,16384,0.15896444850497776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,4096,0.03868444429503547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,3584,0.01166044423977534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,3584,0.03582311007711623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,12288,0.12112088998158772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,3072,0.01054488867521286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,10240,0.10468622048695882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,3072,0.03301689028739929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,2560,0.008054222497675154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,2560,0.029197331931855943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,8192,0.08809866507848103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,2048,0.007172444628344641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,2048,0.02551022171974182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,7168,0.08049244350857206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,1536,0.005967999911970562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,1536,0.022184888521830242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,6144,0.07337155607011583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,1024,0.004921777794758479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,5120,0.06567200024922688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,1024,0.01956266661485036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,768,0.004527111020353106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,768,0.018252443936136033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,4096,0.056177775065104164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,3584,0.05217955509821574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,512,0.004002666721741359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,512,0.01683111157682207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,256,0.0035751110149754416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,256,0.01570666664176517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,3072,0.04920355478922526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,128,0.00342399999499321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,128,0.015231110983424716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,2560,0.04641866683959961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,2048,0.040023111634784274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,64,0.0033448889023727844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,64,0.015048000547620984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,32,0.0033235556135574975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,32,0.015119110544522604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,65536,0.07562844620810615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,65536,0.511224004957411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,16384,0.02680444386270311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,16384,0.13557422161102295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,1536,0.03579377796914842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,12288,0.024881778491867915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,12288,0.09823733568191528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,1024,0.03231911195649041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,10240,0.02275288932853275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,768,0.03149866726663377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,10240,0.0811413327852885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,8192,0.016599110431141324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,8192,0.06719199816385905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,512,0.030237333642111883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,7168,0.014949333336618213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,7168,0.06004710992177328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,256,0.02923111120859782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,6144,0.013456000222100152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,6144,0.05278844303554959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,128,0.029816001653671265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,5120,0.01368711143732071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,5120,0.0464053319560157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,65536,0.5655173195732964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,4096,0.01088177743885252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,4096,0.03939466675122579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,16384,0.15733244684007433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,3584,0.00997866690158844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,3584,0.03572088811132643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,12288,0.11865866184234619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,3072,0.00848711116446389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,3072,0.03199199835459391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,10240,0.10154310862223308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,2560,0.008316444853941599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,8192,0.08686577611499363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,2560,0.02860622273551093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,2048,0.007082666787836287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,2048,0.025082666012975905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,7168,0.07906577984491985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,1536,0.006452444526884291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,1536,0.022126222650210064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,6144,0.07144177622265287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,1024,0.0046826667255825466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,1024,0.0192586663696501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,5120,0.06397600306404962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,768,0.004300444490379757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,768,0.018087110585636564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,4096,0.05536355574925741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,512,0.003863111138343811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,512,0.016900445024172466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,3584,0.05201066533724467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,256,0.0034844444857703317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,256,0.015783111254374187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,3072,0.04744266801410251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,128,0.0032142222755485107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,128,0.015112888481881885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,2560,0.045891556474897593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,64,0.003063999944263034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,64,0.01511111193233066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,2048,0.04074488745795356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,32,0.0031644445326593188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,32,0.015072888798183866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,65536,0.0594080024295383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,65536,0.5112515555487739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,16384,0.02052622205681271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,16384,0.1360364490085178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,1536,0.03616355525122748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,12288,0.019839111301634047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,12288,0.09852355718612671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,1024,0.032088001569112144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,10240,0.018956444329685636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,10240,0.0799751083056132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,768,0.03136266602410211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,8192,0.013600889179441663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,8192,0.06612800227271186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,512,0.030264887544843886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,7168,0.013815999858909182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,256,0.029459555943806965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,7168,0.059860441419813365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,6144,0.013449778159459433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,6144,0.052833778990639575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,128,0.02962755494647556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,5120,0.012269333004951477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,5120,0.04627289043532478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,16384,0.15392533938090006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,4096,0.009852444132169088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,4096,0.03863111138343811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,65536,0.5547573301527235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,3584,0.009015111459626092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,3584,0.03558311197492812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,12288,0.1163911157184177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,3072,0.008163555628723567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,3072,0.03183644347720676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,10240,0.09990044434865315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,2560,0.007279111279381647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,8192,0.08524800009197658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,2560,0.028453333510292902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,2048,0.006646222124497096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,2048,0.02461155586772495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,7168,0.07831911245981853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,6144,0.07060799996058147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,1536,0.0057848890622456866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,1536,0.022163555026054382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,1024,0.004865777575307422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,1024,0.019475556082195707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,5120,0.0621946652730306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,768,0.00435288871328036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,768,0.017996443642510306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,4096,0.05443911088837517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,512,0.004009777886999978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,512,0.016900445024172466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,3584,0.049921777513292104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,256,0.0034808889031410217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,256,0.01572533282968733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,3072,0.047389331791136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,128,0.0032213332338465583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,128,0.015007111761305066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,2560,0.044918222559822925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,64,0.00300177786913183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,64,0.014936000108718872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,32,0.003018666679660479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,32,0.01478666729397244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,65536,0.0514337784714169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,65536,0.5114595625135634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,2048,0.039974222580591835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,16384,0.013826666606797112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,16384,0.13501600424448648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,1536,0.03513600097762214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,1024,0.03177511029773288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,12288,0.012515555653307172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,12288,0.0986257791519165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,10240,0.011239111423492432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,10240,0.0805893341700236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,768,0.030981332063674927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,8192,0.009619555539555019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,8192,0.06590311394797431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,256,0.029045333464940388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,512,0.030461334519916113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,7168,0.011601777540312873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,7168,0.05929866764280531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,6144,0.00903200027015474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,6144,0.05204711026615567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,128,0.02909244431389703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,5120,0.00943022221326828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,5120,0.045828445090187915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,65536,0.5497777726915147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,4096,0.008592888712882996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,4096,0.03792799843682183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,16384,0.152528895272149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,3584,0.007859555383523306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,3584,0.034522665871514216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,12288,0.1156960063510471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,3072,0.007800000409285228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,10240,0.09896088971032037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,3072,0.031140443351533677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,2560,0.006498666687144174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,2560,0.028567999601364136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,8192,0.08386044369803534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,2048,0.0058551111982928375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,2048,0.024718221690919664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,7168,0.07756177584330241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,1536,0.005126222140259213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,1536,0.022055110997623865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,6144,0.06964977582295735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,1024,0.004415110995372136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,1024,0.019408888287014432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,5120,0.062262223826514355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,768,0.004034666551484002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,768,0.017864000466134813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,4096,0.05371555685997009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,512,0.003689777933888965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,512,0.01663377715481652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,3584,0.04993155598640442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,256,0.0033253334048721525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,256,0.015411555767059326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,3072,0.04628088739183214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,128,0.003075555587808291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,128,0.01499555508295695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,2560,0.04292355643378364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,64,0.0029519999192820657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,64,0.0151235560576121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,32,0.002969777832428614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,32,0.014904888139830695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,65536,0.0418124430709415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,65536,0.5114204618665907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,2048,0.037744889656702675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,16384,0.011369778050316704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,16384,0.13633599546220568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,1536,0.034317334493001304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,12288,0.010111110905806223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,12288,0.099016891585456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,1024,0.031657778554492526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,10240,0.010433778166770935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,768,0.030710223648283217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,10240,0.08018844657474093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,8192,0.008582221964995066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,8192,0.06551021999782987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,512,0.030054221550623577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,7168,0.008264000217119852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,7168,0.059395558304256864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,256,0.028569777806599934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,6144,0.009202666580677032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,6144,0.051989333497153394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,128,0.029063112205929224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,5120,0.008611555728647444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,5120,0.045271111859215625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,65536,0.5498577753702799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,4096,0.007822222179836696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,4096,0.03685333331425985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,16384,0.15311288833618164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,3584,0.0075351107451650835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,3584,0.03354933195643955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,12288,0.11680266592237686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,3072,0.007250666618347168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,3072,0.03066133459409078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,10240,0.09935110807418823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,2560,0.006361777583758037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,2560,0.027666666441493567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,8192,0.08442666795518662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,2048,0.00565422202150027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,2048,0.02474133339193132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,7168,0.07765511009428236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,1536,0.004960000101062986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,1536,0.021856889128684998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,6144,0.06966666380564372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,1024,0.00425511101881663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,1024,0.019168888529141743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,5120,0.062124444378746875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,768,0.003946666502290302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,768,0.01791911158296797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,4096,0.05332977904213799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,512,0.003593777616818746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,512,0.016678222351604037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,3584,0.04965155654483371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,256,0.0032622222271230486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,256,0.015458666616015963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,3072,0.04518666532304552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,128,0.003068444422549672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,128,0.015055111712879606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,2560,0.041915555795033775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,64,0.0029537777105967202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,64,0.014765333798196582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,32,0.002955555501911375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,32,0.01459111107720269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,65536,0.035763555102878146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,65536,0.5126746495564779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,16384,0.010429333481523724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,16384,0.1359786722395155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,12288,0.008544888761308458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,12288,0.09839555290010239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,10240,0.007983110845088959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,10240,0.08052266968621148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,8192,0.007430222299363878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,8192,0.06545955604977079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,7168,0.007221333682537079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,7168,0.05968088573879666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,6144,0.006884444504976273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,6144,0.051860445075564914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,5120,0.007344000041484833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,5120,0.04519288738568624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,4096,0.006888000087605582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,2048,0.037800888220469155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,4096,0.0366613335079617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,3584,0.0073626670572492815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,3072,0.007000889215204451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,3584,0.03371466530693902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,2560,0.006208000083764394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,3072,0.030357334348890517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,2048,0.005542222410440445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,2560,0.02758488886886173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,1536,0.004845333182149463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,2048,0.024668445189793903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,1536,0.03462310963206821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,1024,0.0041884444653987885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,1536,0.02177422245343526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,768,0.0038311113086011675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,1024,0.01926488843229082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,512,0.003524444583389494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,768,0.01776622235774994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,256,0.0032008888406885993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,512,0.016536000702116225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,128,0.002987555538614591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,256,0.015339555011855232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,64,0.0028657778683635923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,128,0.01477866702609592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,32,0.002855111120475663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,64,0.014741332994567024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,32,0.014570666684044732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,65536,0.03544444508022732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,16384,0.008234666453467475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,12288,0.007505777809354994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,16384,0.13603199852837458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,10240,0.007270221908887227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,12288,0.09849244356155396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,8192,0.007126222054163615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,65536,0.5126835505167643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,10240,0.08015910784403484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,7168,0.0070462218589252895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,8192,0.06547377506891887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,6144,0.006811555474996567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,7168,0.059375113911098905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,1024,0.03148355417781406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,5120,0.00720177756415473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,6144,0.051873776647779674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,4096,0.006761777732107375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,5120,0.04474311073621114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,3584,0.007239999870459239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,4096,0.037026666932635836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,3072,0.0069191112286514705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,3584,0.03303822212749057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,2560,0.00618577789929178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,3072,0.03015022145377265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,2048,0.005511111269394557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,2560,0.027560000618298847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,1536,0.004861333303981357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,2048,0.02442577812406752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,1024,0.004181333300140169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,1536,0.021740444832377966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,768,0.0038773334688610504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,1024,0.019165333774354722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,768,0.017794667018784415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,512,0.0035075553589397003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,256,0.0031822222388452957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,512,0.016555555992656283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,128,0.002942222274012036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,256,0.01551822159025404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,64,0.0028115556471877625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,128,0.014789332946141561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,32,0.0028524444335036804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,64,0.014941333068741692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,32,0.014576888746685453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,768,0.030807998445298936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,512,0.029899554120169744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,256,0.02885955572128296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,12288,0.8862426545884875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,16384,1.1897111468844943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,10240,0.7382906807793511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,10240,0.6679777569240994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,12288,0.7511839866638184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,16384,0.9943048689100477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,8192,0.49134312735663516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,8192,0.5870889027913412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,128,0.029317332638634577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,7168,0.43994755215115017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,7168,0.5162124633789062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,6144,0.4458026885986328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,6144,0.3799742327796088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,16384,0.6922915776570638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,12288,0.5378959973653158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,5120,0.3700408935546875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,5120,0.31968267758687335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,4096,0.29913422796461314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,4096,0.2615608904096815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,3584,0.2321982118818495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,3584,0.2614106602138943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,10240,0.4520675341288249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,3072,0.22529954380459258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,3072,0.20526133643256295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,8192,0.37747822867499453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,2560,0.1880791054831611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,2560,0.17509867085350883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,7168,0.3379395537906223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,2048,0.14368177784813776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,6144,0.2849280039469401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,2048,0.1539191140068902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,1536,0.1176604429880778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,1536,0.11210044225056966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,1024,0.07959644662009345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,1024,0.08372266425026788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,4096,0.20884266164567736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,5120,0.2472968896230062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,768,0.07244622045093112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,768,0.06604622470008002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,512,0.050880889097849526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,512,0.05287377701865303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,3584,0.19298666053348115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,256,0.03952266772588094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,256,0.042075554529825844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,3072,0.1619431045320299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,128,0.03074666526582506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,128,0.03913333349757724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,2560,0.14569066630469427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,64,0.027240888939963445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,64,0.042466666963365346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,32,0.027959111664030287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,32,0.048531555467181735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,256,0.05871555540296766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,1536,0.10548800230026245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,2048,0.12326488229963516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,512,0.06651822063657972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,16384,0.339812437693278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,768,0.07648444175720215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,128,0.05242844422658285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,1024,0.08366933133867051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,16384,0.23165154457092285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,12288,0.2536097897423638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,12288,0.17806843916575113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,10240,0.21419821845160592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,65536,1.3765618006388347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,10240,0.1519626643922594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,8192,0.1715022193060981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,8192,0.12126044432322185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,7168,0.10863733291625977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,7168,0.15302133560180664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,6144,0.09421777725219727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,6144,0.13365422354804143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,65536,1.0156186421712239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,5120,0.08147378100289239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,5120,0.11234933800167507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,16384,0.2509368790520562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,12288,0.19288978311750624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,4096,0.09302844603856404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,4096,0.06601066721810235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,3584,0.05943377812703451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,3584,0.08296444680955675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,10240,0.16802667246924508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,8192,0.14099821779463026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,3072,0.0520613325966729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,3072,0.07300621933407254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,2560,0.04478489028082954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,65536,0.9619448979695638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,2560,0.06165688567691379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,7168,0.12986489137013754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,2048,0.05249777767393324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,2048,0.036637332704332136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,1536,0.029341333442264136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,1536,0.04327466752794054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,6144,0.11719377835591634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,1024,0.021518222159809534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,1024,0.03421422176890903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,5120,0.1018986635737949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,768,0.017514665921529133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,768,0.030072000291612413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,3584,0.08301511075761583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,512,0.013468444347381592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,512,0.025916443930731878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,4096,0.08667111396789551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,256,0.010072888599501716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,256,0.022600889205932617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,2560,0.06649866369035509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,128,0.007805333369308048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,128,0.019938665959570143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,1536,0.052896890375349254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,64,0.006704000135262807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,64,0.020083554916911654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,32,0.006976000136799282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,32,0.021436444587177698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,3072,0.07496266894870333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,2048,0.05780800183614095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,1024,0.045238223340776235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,768,0.04288444585270352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,256,0.03785955574777391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,16384,0.17240444819132486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,16384,0.2717866632673475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,512,0.040545778142081365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,65536,1.0978871451483834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,128,0.03267911076545715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,12288,0.20326756106482613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,12288,0.1326106654273139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,10240,0.17020089096493193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,10240,0.11119555102454291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,8192,0.09073511097166274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,8192,0.13715644677480063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,7168,0.12221511205037434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,7168,0.08037955231136747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,6144,0.07022222545411852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,6144,0.10823111401663886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,65536,0.6811511251661512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,5120,0.06064889166090223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,5120,0.09106577767266168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,16384,0.23355465465121797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,4096,0.049623999330732554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,65536,0.898164431254069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,4096,0.07599377632141113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,12288,0.18232266108194986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,3584,0.044637332359949745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,3584,0.06788711415396796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,3072,0.03971199856864081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,10240,0.15755289130740696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,3072,0.060673779911465116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,2560,0.034428444173600935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,2560,0.05175822310977512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,8192,0.13128711117638484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,2048,0.02870755394299825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,2048,0.043828444348441235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,7168,0.1192515558666653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,1536,0.023595554961098567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,1536,0.03714044557677375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,6144,0.10718222459157307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,1024,0.016546666622161865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,1024,0.02970577942000495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,5120,0.09194399913152058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,768,0.014067555467287699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,768,0.026334222820070054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,4096,0.080622222688463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,3584,0.0744453337457445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,512,0.011167111496130625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,512,0.024251555403073628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,256,0.00886933339966668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,3072,0.0687733358807034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,256,0.02014222244421641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,128,0.006617777877383762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,128,0.018681777848137748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,2560,0.06270577510197957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,64,0.005773333211739858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,64,0.01880088945229848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,32,0.005960000058015187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,32,0.01910666624704997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,2048,0.05508888761202494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,1536,0.05060177710321215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,1024,0.043565332889556885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,16384,0.15506754981146917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,768,0.04234666625658671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,16384,0.2489217652214898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,65536,1.0049253039889865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,512,0.039776888158586286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,128,0.03314222229851617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,256,0.03573955429924859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,12288,0.18284444014231363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,10240,0.15711732705434164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,12288,0.11981422371334499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,10240,0.10144799947738647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,65536,0.6034711201985677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,8192,0.12705332703060573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,7168,0.11256799432966445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,7168,0.08282666736178927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,8192,0.08190577560000949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,6144,0.09988977511723836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,6144,0.06407466861936781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,5120,0.05485688977771335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,5120,0.08432977729373509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,16384,0.31078489621480304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,65536,1.22708617316352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,4096,0.04508444335725573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,12288,0.24079110887315539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,4096,0.07084977626800537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,3584,0.04085155659251743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,3584,0.06359022193484835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,10240,0.21036711004045275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,8192,0.17400799857245552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,3072,0.036256889502207436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,3072,0.056531555122799344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,2560,0.04758400056097242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,2560,0.031529777579837374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,2048,0.026122666067547266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,7168,0.15502933661142984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,2048,0.04164355662133958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,1536,0.021199999584092036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,1536,0.035192890299691096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,6144,0.13808977603912354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,1024,0.01513777838812934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,1024,0.028340445624457464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,5120,0.1164142158296373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,768,0.012773333324326409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,768,0.02572533322705163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,4096,0.09734311368730332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,3584,0.09110933542251587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,512,0.010481778118345471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,512,0.023056000471115112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,256,0.008002666963471307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,256,0.019447111421161227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,2560,0.06585066848331027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,128,0.00609777785009808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,128,0.018858666221300762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,3072,0.08079022169113159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,64,0.005495999836259418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,64,0.01866222255759769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,32,0.00573333352804184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,32,0.018896889355447557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,1536,0.052173333035575024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,2048,0.058529780970679395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,768,0.041542222102483116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,1024,0.04442044430308872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,16384,0.11843289269341363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,16384,0.20195378197564018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,65536,0.8220524258083768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,12288,0.09077866872151692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,12288,0.15082844098409018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,256,0.03421422176890903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,512,0.03905333413018121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,65536,0.4536782370673285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,128,0.032101333141326904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,10240,0.07717688878377278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,10240,0.1268168952729967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,8192,0.10112088918685913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,8192,0.0628933310508728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,7168,0.055884447362687856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,7168,0.08954044183095296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,6144,0.0489671097861396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,6144,0.08194133308198717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,5120,0.042444444364971586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,5120,0.06929333342446221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,16384,0.17914756139119467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,65536,0.6582008997599283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,4096,0.05905866622924805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,4096,0.034992890225516424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,12288,0.14047733942667642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,3584,0.053485333919525146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,3584,0.031672000885009766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,3072,0.02810933192571004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,3072,0.048624889718161694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,10240,0.12209244569142659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,8192,0.10498755507998997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,2560,0.04134755664401584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,2560,0.024442666106753882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,2048,0.020000000794728596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,2048,0.03545777665244208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,7168,0.1012880007425944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,1536,0.015879111157523263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,1536,0.03052622079849243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,6144,0.08726488881640965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,1024,0.012097777591811286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,1024,0.02552177839808994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,5120,0.0774151086807251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,768,0.010507555471526252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,768,0.023209777143266466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,4096,0.06840088632371691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,512,0.008582221964995066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,3584,0.06624977456198798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,512,0.02092711130777995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,256,0.006322666588756774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,256,0.01793688866827223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,2560,0.051840000682406955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,128,0.0053546664615472155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,128,0.017014222012625802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,768,0.03819288810094198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,64,0.004770666774776247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,64,0.01738666660255856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,32,0.00498933345079422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,32,0.01756000022093455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,1536,0.043506665362252124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,3072,0.058152887556287974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,2048,0.04607111215591431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,1024,0.03838933176464505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,16384,0.19122933016883004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,16384,0.11001244518491958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,65536,0.7664222187466092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,12288,0.14091466532813177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,12288,0.08519199821684097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,10240,0.07330133517583211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,10240,0.11723911762237549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,256,0.03210577699873183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,65536,0.4144497712453206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,128,0.030406223403082952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,8192,0.09394399987326728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,8192,0.056893335448371045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,7168,0.08606221940782334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,7168,0.052912000152799815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,512,0.03568799959288703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,6144,0.04579377836651272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,6144,0.0759342246585422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,5120,0.039424889617496066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,5120,0.06546133094363742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,16384,0.2200088898340861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,4096,0.0328106681505839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,4096,0.05625155236985949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,12288,0.16693600018819174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,65536,0.8485324647691516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,3584,0.051114665137396924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,3584,0.029671112696329754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,3072,0.026455110973782007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,3072,0.04607999987072415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,10240,0.14580889542897543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,2560,0.022815111610624526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,2560,0.03915289044380188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,8192,0.12066845099131267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,2048,0.018590221802393597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,2048,0.03415288858943515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,7168,0.10816444291008843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,1536,0.014870221416155497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,1536,0.029327998558680218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,6144,0.09693066941367255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,1024,0.011762667033407422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,1024,0.02455555564827389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,5120,0.0836720003022088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,768,0.010299555957317352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,768,0.022693332698610093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,3584,0.06908710797627766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,512,0.008657777474986183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,512,0.019852444529533386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,4096,0.07256888680987887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,256,0.006643555644485686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,256,0.017900443739361234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,3072,0.061796446641286217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,128,0.0057546666098965546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,128,0.01703288820054796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,2560,0.05558222532272339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,64,0.0052533331844541765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,64,0.017059556312031217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,32,0.005505777895450592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,32,0.0174346665541331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,1536,0.045532445112864174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,2048,0.04870666729079353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,768,0.03863911165131463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,65536,0.7287608782450358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,16384,0.0998835563659668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,1024,0.04126400086614821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,16384,0.17893689208560518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,12288,0.0785928898387485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,256,0.033201777272754245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,12288,0.13339555263519287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,512,0.03621600071589152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,10240,0.11271111170450847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,128,0.030975110001034204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,10240,0.07863199710845947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,65536,0.3558302190568712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,8192,0.08964444531334771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,8192,0.05079466766781277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,7168,0.0800302227338155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,7168,0.044600887431038745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,6144,0.074054221312205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,6144,0.03856977820396423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,5120,0.03301422132386102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,5120,0.06258400281270345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,65536,0.6276506847805446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,16384,0.1689280006620619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,4096,0.05277422070503235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,4096,0.027775110469924078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,3584,0.025566221939192876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,12288,0.13293066289689806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,3584,0.04917155702908834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,3072,0.022779555784331426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,3072,0.04375733269585503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,10240,0.11478488975101048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,2560,0.019479999939600628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,8192,0.09713421927558051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,2560,0.03808444407251146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,2048,0.01590666671593984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,2048,0.03295822276009454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,7168,0.08942488829294841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,1536,0.014177777700954013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,1536,0.02860800094074673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,6144,0.08080266581641303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,1024,0.010070222119490305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,1024,0.024636444118287828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,5120,0.07145955827501085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,768,0.008814222282833522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,768,0.021542222963439092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,4096,0.06362577941682604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,512,0.0074311114019817775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,512,0.01964977714750502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,3584,0.061384889814588756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,256,0.005622222191757626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,256,0.017856889300876193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,2560,0.05089599887530009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,128,0.0047182221379545005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,3072,0.055289778444502086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,128,0.01703288820054796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,64,0.0042204442951414315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,64,0.01696888936890496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,32,0.004481777962711123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,2048,0.04502222273084852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,32,0.017091555727852713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,1536,0.04297599858707852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,768,0.03661777906947666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,1024,0.03822222352027893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,16384,0.08391910791397095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,16384,0.17108799351586235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,65536,0.6995262040032281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,12288,0.1274720033009847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,12288,0.06596977843178643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,10240,0.05642933315700955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,10240,0.10656177997589111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,65536,0.3194053437974718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,8192,0.04389511214362251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,512,0.035467555125554405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,256,0.03189333279927572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,8192,0.08816444211535984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,7168,0.07897777689827813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,128,0.030200888713200886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,7168,0.040185779333114624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,6144,0.07063821951548259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,6144,0.0350195566813151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,5120,0.030689779255125258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,5120,0.06047999858856201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,16384,0.20771288871765137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,4096,0.025475554996066626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,4096,0.05191644363933139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,65536,0.7854071193271213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,12288,0.16093599796295166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,3584,0.025444444682863023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,3584,0.04788444439570109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,3072,0.02016888890001509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,3072,0.04243377844492594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,10240,0.14020355542500815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,2560,0.016943999462657504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,2560,0.03651377889845107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,8192,0.11370044284396702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,2048,0.013719999955760108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,2048,0.03220266766018338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,7168,0.10496266682942708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,1536,0.011697777443461947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,1536,0.027905778752432928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,6144,0.09389777978261311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,1024,0.009268444445398118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,5120,0.08122399780485365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,1024,0.023591111103693645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,768,0.008106666306654612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,768,0.020819556381967332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,3584,0.06618133518430921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,512,0.007050666544172499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,512,0.01908533275127411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,4096,0.07002666923734877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,256,0.005106666849719154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,256,0.017440888616773818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,3072,0.058966219425201416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,128,0.004559999952713649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,128,0.016771554946899414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,2560,0.05134400063090854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,64,0.0042133331298828125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,64,0.016819554898473952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,32,0.004314666820896996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,32,0.016938666502634685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,1536,0.042669332689709134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,2048,0.0458568897512224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,768,0.03667555583847894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,1024,0.03793511125776503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,65536,0.5514497756958008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,65536,0.24675022231207955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,16384,0.09438577625486587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,16384,0.13410400019751653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,12288,0.09948266877068414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,12288,0.07585866583718194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,10240,0.067193779680464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,10240,0.08443378077612983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,256,0.031974222924974226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,512,0.03331644336382548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,8192,0.06863911284340753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,128,0.030304888884226482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,8192,0.049330668316947095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,7168,0.0628017783164978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,7168,0.045384890503353544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,6144,0.05760000149408976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,6144,0.03961866762903001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,5120,0.03400889039039612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,5120,0.04993866549597847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,16384,0.15432622697618273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,65536,0.5655297703213161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,4096,0.04324000080426534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,4096,0.020410666863123577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,3584,0.018773333893881906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,3584,0.0392231113380856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,10240,0.10660977496041192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,12288,0.1213759978612264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,3072,0.01735466718673706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,3072,0.036129779285854764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,8192,0.09141066339280869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,2560,0.014697778556081982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,2560,0.031065778599845037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,2048,0.012560000022252401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,2048,0.028043554888831243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,6144,0.0763546692000495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,1536,0.010233778092596266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,1536,0.02500444485081567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,7168,0.08400355445014106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,1024,0.00793511089351442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,1024,0.020615110794703167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,5120,0.06796355379952325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,768,0.006962666908899943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,768,0.019056888090239633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,3584,0.05702755848566691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,512,0.005433777968088786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,512,0.01763911048571269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,768,0.033067554235458374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,256,0.004600000050332812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,256,0.016375111209021676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,2560,0.04682755470275879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,128,0.004023111114899318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,128,0.015754666593339708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,4096,0.06003733476003011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,64,0.003687111039956411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,64,0.0159244438012441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,32,0.0038853333228164246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,32,0.015824000040690105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,1536,0.0402124457889133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,2048,0.04187022315131294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,3072,0.0513591104083591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,1024,0.03583911061286926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,65536,0.5263822343614366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,16384,0.0626151098145379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,16384,0.1272746721903483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,12288,0.0492266681459215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,12288,0.09458755784564549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,10240,0.04519911275969612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,65536,0.22618754704793295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,10240,0.08037244611316256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,8192,0.033844444486829973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,8192,0.06610577636294894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,128,0.029263112280103896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,7168,0.030961778428819444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,7168,0.06059555874930488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,256,0.031189332405726116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,6144,0.02966488732231988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,6144,0.053370667828453906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,512,0.032116442918777466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,5120,0.02365066607793172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,5120,0.04773955543835958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,16384,0.15124444166819254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,4096,0.01920622256067064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,65536,0.5494497617085775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,4096,0.04136711027887132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,3584,0.01702311138312022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,3584,0.037920888927247785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,12288,0.11950933933258057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,3072,0.014924445086055331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,3072,0.034725334909227155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,10240,0.10453244712617661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,2560,0.01406044430202908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,2560,0.030304001437293157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,8192,0.08937422434488933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,2048,0.011143111520343356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,2048,0.02730488942729102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,7168,0.0830213361316257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,1536,0.011445333560307821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,1536,0.024433778391944036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,6144,0.07502844598558214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,1024,0.008999110923873054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,1024,0.020206222931543987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,5120,0.06674133406745063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,768,0.007782222496138678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,768,0.018817777434984844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,3584,0.05585155884424845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,512,0.0058062221441004015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,512,0.017481777403089736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,2560,0.04698400033844841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,256,0.0047564444442590075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,3072,0.05097155438529121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,256,0.016744888491100736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,128,0.004250666747490565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,4096,0.058428446451822914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,128,0.01590577761332194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,64,0.003951999876234266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,64,0.015783111254374187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,32,0.004179555508825514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,32,0.01590755581855774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,2048,0.04197600152757433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,65536,0.49593687057495117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,1536,0.040022220876481794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,65536,0.19113511509365508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,16384,0.054231110546323985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,16384,0.11584266026814778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,12288,0.04227911101447212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,12288,0.08763021892971462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,1024,0.03370577759212918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,768,0.03247377938694424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,10240,0.07621510823567708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,10240,0.038988444540235735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,8192,0.029488888051774766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,8192,0.06394044558207194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,128,0.028929776615566675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,7168,0.028661333852344092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,7168,0.05861777729458279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,256,0.030630220969518025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,6144,0.02385155525472429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,6144,0.053120000494851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,512,0.03214222192764282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,5120,0.021355556117163763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,5120,0.04555199874771965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,16384,0.14570666684044734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,4096,0.016252444850073922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,4096,0.0392542216512892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,65536,0.5347519980536567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,3584,0.016417778200573392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,3584,0.03637689020898607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,12288,0.11266222265031602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,3072,0.013068444199032254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,10240,0.09880355331632827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,3072,0.03275555703375075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,2560,0.01315199997689989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,2560,0.02887822190920512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,8192,0.08472177717420791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,2048,0.011301333705584208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,2048,0.026571555270089045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,7168,0.07810399929682414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,1536,0.0092604441775216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,1536,0.02311022248533037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,6144,0.07163911395602755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,1024,0.007428444094128079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,1024,0.019914666811625164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,5120,0.06373600165049235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,768,0.006114666660626729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,768,0.018699554933442008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,4096,0.056903998057047524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,512,0.0052871112194326185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,512,0.017551110850440133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,3584,0.054486221737331815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,256,0.00424888895617591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,256,0.01630311045381758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,3072,0.05049066742261251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,128,0.0037973332736227247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,128,0.015614221493403116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,2560,0.046386665768093534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,64,0.0036391110883818734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,64,0.015615999698638916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,32,0.0038204445607132385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,2048,0.04142399960094028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,32,0.015424888994958667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,1536,0.03991644581158956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,768,0.03274755676587423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,65536,0.4317928949991862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,65536,0.16767199834187826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,1024,0.03397599856058756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,16384,0.052686221069759794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,16384,0.1074720025062561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,12288,0.08532800277074178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,12288,0.038346668084462486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,10240,0.034323556555642024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,10240,0.07368977864583333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,8192,0.026800000005298193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,8192,0.06158044603135851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,512,0.03196355700492859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,256,0.030789332257376775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,7168,0.055323554409874805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,7168,0.02420266634888119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,6144,0.021359999974568684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,128,0.028912888632880315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,6144,0.05090044604407417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,5120,0.01789955629242791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,5120,0.043354666895336576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,16384,0.13531466325124106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,12288,0.10635377963383992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,4096,0.014163555370436775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,4096,0.03625244564480252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,3584,0.012852444416946836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,65536,0.4983528984917535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,10240,0.09308622280756633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,3584,0.03398488958676656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,3072,0.012675555215941535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,3072,0.031389332479900785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,2560,0.011183111204041375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,2560,0.02780710988574558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,8192,0.08125422398249309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,2048,0.008965333302815756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,2048,0.025941333836979334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,7168,0.07443288962046306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,5120,0.06100444661246406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,6144,0.0682577755716112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,1536,0.008098666866620382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,1024,0.006179555422729916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,1536,0.022069333328141108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,1024,0.01953866746690538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,768,0.005234666582610872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,768,0.018427555759747822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,3584,0.05174577898449368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,512,0.004716444346639845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,512,0.017380444539917838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,4096,0.053483555714289345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,3072,0.04773599902788798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,256,0.003961777935425441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,256,0.016283555163277518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,128,0.003552888830502828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,128,0.015446222490734525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,2560,0.044350223408805005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,64,0.0035617777870761026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,64,0.015483554866578845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,32,0.003572444534964032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,32,0.01534400052494473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,2048,0.039805332819620766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,65536,0.4516604211595323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,65536,0.13273421923319498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,16384,0.05445955528153313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,768,0.03185866607560052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,16384,0.10615022314919366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,1536,0.038409776157803006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,12288,0.08459999826219346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,1024,0.03242222136921353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,12288,0.04757510953479343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,10240,0.06948266426722209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,10240,0.04145155681504144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,8192,0.03404266635576884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,8192,0.060805333985222705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,128,0.028460443019866943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,7168,0.029000888268152874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,256,0.03070311082734002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,7168,0.05499200026194254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,6144,0.02535733415020837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,512,0.031155556440353394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,6144,0.04973511232270134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,5120,0.019204444355434842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,5120,0.04229066769282023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,65536,0.505533324347602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,4096,0.01389955480893453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,4096,0.0365591115421719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,16384,0.13918756114112005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,3584,0.013016888664828407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,3584,0.03417422374089559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,12288,0.10896622472339207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,3072,0.01238044434123569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,3072,0.031848000155554876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,10240,0.09575644466612075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,2560,0.01070666644308302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,2560,0.02776088813940684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,8192,0.08240444130367704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,2048,0.008919110728634728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,2048,0.02546399997340308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,7168,0.07547466622458564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,1536,0.007526222202512953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,1536,0.021990221407678392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,6144,0.0684995585017734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,1024,0.0058408888677755995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,1024,0.019433778193261888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,5120,0.06074577569961548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,768,0.005130666825506422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,768,0.018352000249756705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,4096,0.05393422312206692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,512,0.004368888835112254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,512,0.017280000779363845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,3584,0.05106311043103536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,256,0.003891555385457145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,256,0.016191111670600045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,3072,0.04749155706829495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,128,0.0035537779331207275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,2560,0.04354755414856804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,128,0.015480000111791821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,64,0.0033039999090962936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,64,0.01553066737122006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,32,0.0034382223255104492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,32,0.015228443675571017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,2048,0.038966221941841975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,65536,0.13535555203755698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,768,0.031017776992585924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,16384,0.036786665519078575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,16384,0.10663733217451309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,65536,0.4281760056813558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,1536,0.036362667878468834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,12288,0.08132177591323853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,1024,0.03147111005253262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,12288,0.03057066599527995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,10240,0.025793777571784124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,10240,0.06990844673580594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,8192,0.021039111746682063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,8192,0.058674666616651744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,256,0.03024622135692173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,7168,0.0189057770702574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,7168,0.05345955491065979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,128,0.02868710954984029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,6144,0.017095999585257638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,6144,0.04675111174583435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,512,0.030660443835788306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,5120,0.014475555883513557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,5120,0.04083555605676439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,16384,0.1289502249823676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,4096,0.012164444559150271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,4096,0.035813334915373064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,65536,0.48763110902574325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,3584,0.011355555719799466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,12288,0.1007688906457689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,3584,0.03356533249219259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,3072,0.010026666853162978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,10240,0.08956355518764919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,3072,0.03128977616628011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,2560,0.010255999863147736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,2560,0.026731555660565693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,8192,0.07756977611117892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,2048,0.009352000223265754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,2048,0.025127111209763422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,7168,0.07160088751051161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,1536,0.0075662218862109715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,6144,0.065138664510515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,1536,0.021771555145581562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,1024,0.00563466673096021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,1024,0.019509333703253005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,5120,0.05746044715245565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,768,0.00498577786816491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,768,0.018076444665590923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,3584,0.049525333775414355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,512,0.004382222063011593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,512,0.017120000388887193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,4096,0.05057422320048014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,256,0.0038088887102074097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,256,0.01610488858487871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,3072,0.046421332491768726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,128,0.0034604445099830627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,128,0.015263110399246216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,2560,0.04280266828007168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,64,0.0033688888781600525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,64,0.01536266671286689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,32,0.003293333368168937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,32,0.01512888901763492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,2048,0.038014223178227745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,65536,0.425542222128974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,65536,0.10435911019643147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,16384,0.0292871097723643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,16384,0.10670489072799683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,1536,0.035607109467188515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,1024,0.0322746667597029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,12288,0.027820444769329492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,12288,0.07982666624916925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,768,0.031339555978775024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,10240,0.024884444144037034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,10240,0.06898844242095947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,8192,0.01682933337158627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,8192,0.058145774735344775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,512,0.03096533483929104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,256,0.02980177932315403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,128,0.028151111470328435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,7168,0.01497333414024777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,7168,0.052707556221220225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,6144,0.013436444103717804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,6144,0.04693333307902018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,5120,0.011937778029176923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,5120,0.04009688893953959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,65536,0.45357513427734375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,4096,0.010219555762079027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,16384,0.12238578001658122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,4096,0.03541422221395705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,3584,0.009444444543785518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,3584,0.03324710991647508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,12288,0.09637066390779282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,3072,0.008740444150235917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,3072,0.02973244587580363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,10240,0.08454666535059611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,2560,0.008763555851247575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,2560,0.027239110734727647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,8192,0.07362044519848294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,2048,0.006831999868154526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,2048,0.023745778534147475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,7168,0.06751200225618151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,1536,0.006005333529578314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,6144,0.061511112584008105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,1536,0.021384000778198242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,1024,0.004819555415047539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,1024,0.019031110737058852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,5120,0.05473777651786804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,768,0.00434044458799892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,768,0.017886221408843994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,4096,0.04910666743914286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,3584,0.047482666042115956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,512,0.003999999827808804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,512,0.016884444488419425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,256,0.0035617777870761026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,256,0.01574577722284529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,3072,0.04496177699830797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,128,0.0033617777129014335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,128,0.015154666370815702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,2560,0.040954666005240545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,64,0.003157333367400699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,64,0.01513066722287072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,32,0.003257777748836411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,2048,0.035424888134002686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,32,0.015088889333936902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,65536,0.4239528973897298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,65536,0.06720444228914049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,16384,0.02515911062558492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,16384,0.10709155268139309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,1536,0.0351528889603085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,12288,0.021176000436147053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,1024,0.030850665436850652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,12288,0.08088711235258314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,10240,0.019512888458040025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,10240,0.06856622298558553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,768,0.030229333374235366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,8192,0.016168889072206285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,8192,0.05762933360205757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,512,0.030128889613681372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,7168,0.014853333433469137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,7168,0.05206400156021118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,256,0.029992000924216375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,6144,0.013180444637934366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,6144,0.046422223250071205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,128,0.02833777666091919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,5120,0.011611555185582904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,5120,0.03991733325852288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,16384,0.12141156196594238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,4096,0.010032888915803697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,4096,0.035086221165127225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,65536,0.44397687911987305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,3584,0.009248889154858058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,3584,0.03309955530696445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,12288,0.0958862238460117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,3072,0.008453333543406593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,3072,0.02975644336806403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,10240,0.08472800254821777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,2560,0.007430222299363878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,2560,0.025807110799683466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,8192,0.07340621948242188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,2048,0.006513777706358168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,2048,0.023489778240521748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,7168,0.06738311052322388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,1536,0.005776888794369168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,1536,0.02124622298611535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,6144,0.061412447028689914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,1024,0.0047688889834615915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,1024,0.018746667438083224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,5120,0.05452000101407369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,768,0.004369777937730153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,768,0.01776444415251414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,4096,0.04822488956981235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,512,0.0038924444880750445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,512,0.01661066710948944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,3584,0.04742133286264208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,256,0.003458666718668408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,256,0.015764445066452026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,3072,0.044518222411473594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,128,0.003237333355678452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,2560,0.040291554398006864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,128,0.01497866710027059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,64,0.003089777711364958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,2048,0.034953776333067156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,32,0.0031271111220121384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,64,0.014916444818178812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,32,0.014898666077189975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,65536,0.4247093465593126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,65536,0.05351110961702135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,1536,0.034070223569869995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,16384,0.01756533318095737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,16384,0.1067697803179423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,12288,0.016307555966907077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,12288,0.07941689093907674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,1024,0.03074666526582506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,10240,0.014670222997665405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,10240,0.06864621904161242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,768,0.03108888864517212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,8192,0.013049778011110095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,8192,0.05661333269543118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,512,0.030332446098327637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,7168,0.012281777958075205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,7168,0.051826665798823036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,256,0.030045333835813735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,6144,0.012175111307038201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,6144,0.0459591117170122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,128,0.02795999911096361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,5120,0.010841777755154504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,5120,0.0401777790652381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,16384,0.1195146640141805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,4096,0.00924444446961085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,4096,0.03446222345034281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,12288,0.09379911422729492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,3584,0.008092444803979661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,65536,0.43486666679382324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,3584,0.03281955586539374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,3072,0.007672888537247975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,3072,0.028825776444541082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,10240,0.08293066422144572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,2560,0.007540444533030192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,2560,0.02607911162906223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,8192,0.07208533419503106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,2048,0.007285333342022366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,2048,0.023296000228987798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,7168,0.06617599725723267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,1536,0.0058044443527857465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,1536,0.021112889051437378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,6144,0.0596639977561103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,1024,0.004893333133724001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,1024,0.019072888625992667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,5120,0.05322666631804573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,768,0.004371555729044808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,768,0.017638223038779367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,4096,0.04761688907941183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,512,0.003863111138343811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,512,0.016606221596399944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,3072,0.041851556963390775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,3584,0.04587022132343716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,256,0.003440888805521859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,256,0.015612444943851896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,128,0.0031822222388452957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,128,0.015020444989204407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,2560,0.039184000757005476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,64,0.002969777832428614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,64,0.015000889698664347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,32,0.002992000016901228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,32,0.014880888991885714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,65536,0.04673422376314799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,65536,0.42534756660461426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,2048,0.03482044405407376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,16384,0.01198933356338077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,16384,0.1064079999923706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,1536,0.03373688790533278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,12288,0.012684444586435953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,12288,0.07931911283069186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,1024,0.030802667140960693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,10240,0.011687111523416309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,10240,0.06825688812467787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,768,0.03003822101487054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,8192,0.010050666828950247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,8192,0.05681777662701077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,512,0.030251555972629126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,7168,0.009511111511124505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,256,0.02946044339074029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,7168,0.05131733417510986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,6144,0.008939555949634975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,6144,0.04586222105556064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,128,0.02778488728735182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,5120,0.009099555512269339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,5120,0.03919822308752272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,65536,0.43074666129218203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,4096,0.008038221961922115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,4096,0.03372088736957974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,16384,0.1178942256503635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,3584,0.007751999629868402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,3584,0.030404445197847154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,12288,0.09295644362767537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,3072,0.007044444481531779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,3072,0.02847555610868666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,10240,0.08180266618728638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,8192,0.07024977604548137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,2560,0.006384000182151794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,2560,0.0255404445860121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,2048,0.005695999910434087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,2048,0.023423999547958374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,7168,0.06503555509779188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,1536,0.005037333402368757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,6144,0.059564444753858775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,1536,0.021207110749350652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,1024,0.004318222403526306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,1024,0.018887110882335238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,5120,0.05242844422658285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,768,0.003960000144110786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,768,0.0177848885456721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,4096,0.045175111956066556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,512,0.0036115555299652945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,512,0.016534222496880423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,3584,0.0444088876247406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,256,0.0033048888047536216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,256,0.015572445260153877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,3072,0.040489776266945735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,128,0.0030542222989930045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,128,0.014817777607176038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,2560,0.037672887245814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,64,0.002899555489420891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,64,0.014835556348164877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,32,0.0029102222373088202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,32,0.01461155547036065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,65536,0.037652442852656044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,65536,0.4257697794172499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,2048,0.03438755538728502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,16384,0.013443555268976422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,16384,0.10700622532102798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,1536,0.033555554019080266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,12288,0.010998222563001843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,12288,0.07991111278533936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,1024,0.03048622277047899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,10240,0.009922666682137383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,10240,0.06825600067774455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,768,0.029947555727428857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,8192,0.00886399961180157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,8192,0.05651466714011299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,512,0.030250665214326646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,7168,0.008434666527642144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,7168,0.05060177710321215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,256,0.029545777373843726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,6144,0.00813244448767768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,6144,0.04589955674277412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,128,0.027699554959932964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,5120,0.0074728892909155945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,5120,0.03856711255179511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,65536,0.4301057656606038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,4096,0.007052444749408298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,4096,0.032467554012934365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,16384,0.11707911226484512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,3584,0.007483555210961237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,3584,0.0304097765021854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,12288,0.09265955289204915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,3072,0.007106666763623555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,3072,0.028206222587161597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,10240,0.08177955283059014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,2560,0.006383111079533895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,8192,0.0703386664390564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,2560,0.025438222620222304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,2048,0.005678222411208683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,2048,0.02308533257908291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,7168,0.06446844339370728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,1536,0.004998222407367495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,1536,0.020766221814685397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,6144,0.05905866622924805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,1024,0.004256000121434529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,1024,0.01849244369400872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,5120,0.05232355660862393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,768,0.003924444317817688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,768,0.017236444685194228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,4096,0.04589244392183092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,512,0.003547555456558863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,512,0.016143111719025504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,3584,0.0433208876185947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,256,0.0032515554792351196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,256,0.015435554915004306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,3072,0.040848887628979154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,128,0.003023111157947116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,128,0.014764444695578681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,2560,0.03789066606097751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,64,0.0028577778074476454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,64,0.014646222194035849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,32,0.00286666676402092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,2048,0.0339626669883728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,32,0.014578666951921252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,65536,0.030755556292004053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,16384,0.01092266705301073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,16384,0.10711644755469428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,65536,0.4260479874081082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,12288,0.009463110731707679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,10240,0.008792888787057664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,12288,0.08043466673956977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,10240,0.06821778085496691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,8192,0.008016889293988546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,7168,0.007368889119890001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,8192,0.05695377455817329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,7168,0.050847109821107656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,6144,0.0069431112044387394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,6144,0.0450835559103224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,5120,0.007370666497283512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,5120,0.03769866625467936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,4096,0.006853333363930385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,4096,0.0326275560590956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,3584,0.007324444750944774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,3584,0.030400001340442236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,3072,0.006970666348934174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,3072,0.028359111812379625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,2560,0.006231110956933763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,1536,0.033632887734307185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,2560,0.025417778227064345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,2048,0.005580444302823808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,2048,0.02310755517747667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,1536,0.004913777940803104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,1024,0.004234666625658671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,1536,0.020865778128306072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,768,0.0038897775941424896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,1024,0.018505778577592637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,768,0.017408889200952318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,512,0.003552888830502828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,256,0.003248888792263137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,512,0.01623911162217458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,128,0.0030008889734745026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,256,0.015431111057599386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,64,0.0028728888266616394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,128,0.01482311056719886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,32,0.0028293333533737394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,64,0.014696000350846184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,32,0.01458044515715705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,65536,0.028543998797734577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,16384,0.00903911143541336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,65536,0.4263377719455295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,12288,0.008254222571849823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,16384,0.10718488693237305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,10240,0.007471111085679796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,12288,0.0802604423628913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,8192,0.007370666497283512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,10240,0.06795466608471341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,7168,0.0069475554757648045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,8192,0.056624001926845975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,6144,0.006781333436568578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,7168,0.05040622088644239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,5120,0.007167999943097432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,6144,0.044176889790429004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,4096,0.006822222222884496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,5120,0.037668443388409086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,3584,0.007224889265166388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,4096,0.03271289004219903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,3072,0.0070168889231152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,3584,0.03035199973318312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,2560,0.0063075555695427795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,3072,0.028326223293940227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,2048,0.005541333307822545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,2560,0.02539377742343479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,1536,0.004867555366622077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,2048,0.023079999619060095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,1024,0.004273777620659934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,1536,0.020780444145202637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,768,0.0038933331767717996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,1024,0.018447111050287884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,512,0.0035866668654812705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,768,0.017394666870435078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,256,0.0031804444475306403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,512,0.01620444489849938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,128,0.003136000078585413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,256,0.015263999501864115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,128,0.014793778459231058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,64,0.0028106667515304354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,32,0.002830222249031067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,64,0.014597333139843412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,1024,0.030635555585225422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,32,0.014665777484575907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,768,0.02968977888425191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,512,0.02993688980738322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,256,0.029358221424950495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,128,0.027813331948386297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,12288,0.36217154396904844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,16384,0.4636551009284125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,10240,0.30822134017944336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,8192,0.24833689795600045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,12288,0.6100924279954698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,8192,0.4077813360426161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,10240,0.5050364600287544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,7168,0.22946667671203613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,16384,0.7877475420633951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,7168,0.36224712265862363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,12288,0.42909955978393555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,6144,0.21585867140028211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,16384,0.5256204605102539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,6144,0.31590843200683594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,5120,0.1654533280266656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,10240,0.368106656604343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,5120,0.2659297784169515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,4096,0.2176053391562568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,4096,0.13380444049835205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,3584,0.13408354918162027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,8192,0.2978115611606174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,3584,0.19483555687798393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,3072,0.10645244518915813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,3072,0.17052088843451607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,7168,0.2682880030737983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,2560,0.09335555632909139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,2560,0.14528977870941162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,2048,0.07894666989644368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,2048,0.11888710657755534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,1536,0.09394044346279568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,1536,0.06629155741797553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,6144,0.23788089222378203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,1024,0.06621066729227702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,5120,0.1962240007188585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,1024,0.05392533209588793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,4096,0.16842756006452772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,768,0.04696977800793118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,768,0.0536373323864407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,3584,0.147944000032213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,512,0.04150044586923387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,512,0.041080888774659895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,256,0.0295448899269104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,256,0.033430222007963396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,3072,0.13164177205827501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,128,0.022113778524928625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,128,0.02871111035346985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,2560,0.11674666404724121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,64,0.016799110505315993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,64,0.03056266572740343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,32,0.016672000288963318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,32,0.035275555319256253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,2048,0.10072888930638631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,1536,0.08509066369798447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,256,0.04925688770082262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,512,0.05669866667853462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,768,0.0646524429321289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,128,0.04292000002331204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,16384,0.15570488240983751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,1024,0.07064000103208753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,16384,0.20189155472649467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,12288,0.12390932771894668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,12288,0.1570311122470432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,65536,0.5930053393046061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,10240,0.10757866832945083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,10240,0.1309537755118476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,8192,0.088191999329461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,8192,0.10642844438552856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,7168,0.08242844210730658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,7168,0.09440622064802383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,6144,0.08259644773271349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,6144,0.07111110952165392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,65536,0.7807404200236002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,5120,0.06970310873455472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,5120,0.06848355796602038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,16384,0.1902195612589518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,4096,0.057855996820661754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,4096,0.05449688765737745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,12288,0.1518471108542548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,10240,0.13405333624945745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,3584,0.05184088812934028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,3584,0.04715377754635281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,65536,0.7011742062038846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,3072,0.04593244526121351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,3072,0.04269688990381029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,8192,0.11826399962107341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,2560,0.03951822386847602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,2560,0.03855999973085191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,2048,0.032948444286982216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,7168,0.10214666525522868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,2048,0.033844444486829973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,1536,0.02664622167746226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,1536,0.029439111550649006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,6144,0.09269155396355523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,1024,0.018630221486091614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,1024,0.025193777349260118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,5120,0.08474489053090413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,768,0.014690667390823364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,768,0.022222222553359136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,3584,0.06711377700169881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,512,0.011117333339320289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,512,0.021056888831986323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,4096,0.07126488950517443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,256,0.00853777759604984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,256,0.017708443933063082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,2560,0.05674311187532213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,128,0.006183111005359226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,128,0.016725333200560678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,3072,0.060864885648091636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,64,0.005399111244413588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,64,0.01723377737734053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,32,0.005737777799367905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,32,0.017717333303557504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,1536,0.046442667643229164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,2048,0.05137155453364054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,768,0.04083822170893351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,1024,0.04178222351604038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,256,0.03490399983194139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,512,0.03734133309788174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,65536,0.5510329140557183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,128,0.030621333254708186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,16384,0.14440533849928114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,12288,0.11429866154988606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,12288,0.11609778139326309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,16384,0.15018578370412192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,10240,0.09952711396747166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,10240,0.09706311093436347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,8192,0.08128088712692261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,65536,0.570634682973226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,8192,0.07882577843136258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,7168,0.07087733348210652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,7168,0.0784426662656996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,6144,0.06643378072314791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,6144,0.06207288636101616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,5120,0.05283022258016798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,5120,0.058394663863711886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,16384,0.16333156161838108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,65536,0.6043733490837945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,4096,0.05016711023118761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,4096,0.04361777835422092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,12288,0.13175822628868952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,3584,0.03917955689960056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,3584,0.043440000878440015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,3072,0.03480266531308492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,10240,0.11590577496422662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,8192,0.10343822505738999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,3072,0.03895733422703213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,2560,0.035084442959891424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,2560,0.030018667380015056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,2048,0.025044444534513686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,2048,0.030954665607876245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,7168,0.08948177761501735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,1536,0.020055111911561754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,6144,0.08316089047325982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,1536,0.027239110734727647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,1024,0.01384622189733717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,1024,0.02327466673321194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,5120,0.0753600001335144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,768,0.01143022212717268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,768,0.022170666191313002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,3584,0.06009511152903239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,512,0.009205333060688442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,512,0.020468443632125854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,4096,0.06337866518232557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,256,0.007267555428875818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,256,0.017457778255144756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,2560,0.051033778323067554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,128,0.0052471111218134565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,128,0.016774222254753113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,3072,0.05493333273463779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,64,0.0046897778908411665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,64,0.017055999901559617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,32,0.005034666508436203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,32,0.017440888616773818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,2048,0.046351999044418335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,768,0.03827288746833801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,1536,0.04266488883230421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,1024,0.03897422220971849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,16384,0.12862133979797363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,16384,0.1415991120868259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,65536,0.5424070888095432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,12288,0.10018666585286458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,256,0.03204266561402215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,12288,0.11282577779557969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,128,0.02977422210905287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,10240,0.09753421942392985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,512,0.03537599907981025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,10240,0.08389511373307969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,8192,0.07935466368993123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,8192,0.06824533144632976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,7168,0.06086577971776327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,7168,0.07209066549936931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,65536,0.5366071065266927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,6144,0.05379111236996121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,6144,0.06864355670081244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,5120,0.04621244470278422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,5120,0.057803557978736035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,16384,0.15319466590881348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,4096,0.03784177700678507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,4096,0.04615644282764859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,65536,0.550338692135281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,3584,0.034344888395733304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,3584,0.042542222473356456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,12288,0.1235555542839898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,10240,0.10905510849422878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,3072,0.03348177671432495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,3072,0.037957333856158786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,2560,0.02651111119323307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,2560,0.03598577777544657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,8192,0.09310577975379096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,2048,0.022112000319692824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,2048,0.03033155534002516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,7168,0.08542578087912665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,1536,0.017403556240929496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,1536,0.026138666603300307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,6144,0.07948178052902222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,1024,0.012568000290128918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,1024,0.02299822204642826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,5120,0.07245688968234591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,768,0.01090222183201048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,768,0.021873777111371357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,4096,0.060644441180759005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,512,0.008944888909657797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,512,0.018704000446531508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,3584,0.0580000016424391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,256,0.006866666591829724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,256,0.01732711162832048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,2560,0.048617776897218495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,128,0.005160000175237656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,128,0.016629333297411602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,3072,0.052779555320739746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,64,0.0046986668474144405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,64,0.01683200067943997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,32,0.00488711107108328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,32,0.017216889394654166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,1536,0.040561778677834406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,2048,0.04419999983575609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,768,0.03756355577045017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,1024,0.03808355662557814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,16384,0.10347911384370591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,16384,0.10473155313067967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,65536,0.39402132564120823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,12288,0.08446044392055935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,12288,0.08104533619350858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,10240,0.06797155406739977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,128,0.029492444462246362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,256,0.03144711256027222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,10240,0.07456088728374906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,65536,0.41675734519958496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,8192,0.06163555383682251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,8192,0.054697778489854604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,7168,0.05745689074198405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,7168,0.049313777022891574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,6144,0.043940444787343345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,512,0.03457155492570665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,6144,0.052130666044023305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,5120,0.037393778562545776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,5120,0.04638488756285774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,16384,0.1388746632470025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,12288,0.11296178234948052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,65536,0.48562664455837673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,4096,0.03087466624048021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,4096,0.03719110952483283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,3584,0.03421599997414483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,3584,0.028006222512986925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,3072,0.02493600050608317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,3072,0.03145066565937466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,10240,0.09887111186981201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,2560,0.02123822271823883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,2560,0.028765334023369685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,8192,0.08810755279329087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,2048,0.017437333861986797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,2048,0.02552177839808994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,7168,0.07807022333145142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,1536,0.01332800007528729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,1536,0.024002666274706524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,6144,0.07201244433720906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,1024,0.010212444596820408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,1024,0.021185777253574792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,5120,0.06561244196361966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,768,0.008749333520730337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,768,0.01887911061445872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,4096,0.05603555838267008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,512,0.007227555745177799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,3584,0.0525875555144416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,512,0.017282666431532968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,256,0.005210666606823604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,256,0.01611199975013733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,3072,0.04816710948944092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,128,0.00462311092350218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,128,0.01588622232278188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,2560,0.04496177699830797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,64,0.004053333153327306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,64,0.015628443823920358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,32,0.004374222209056218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,32,0.015860444969601102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,2048,0.0410773323641883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,1536,0.038712001509136625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,1024,0.03604177633921305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,768,0.03648711244265238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,16384,0.09001244439019097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,65536,0.37830932935078937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,16384,0.10027199983596802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,12288,0.08115466435750325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,512,0.03299199872546726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,256,0.0306311117278205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,12288,0.07056622372733222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,10240,0.07125600179036458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,128,0.02919644448492262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,10240,0.05878844526078966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,8192,0.05870311127768623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,8192,0.04823644293679131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,7168,0.054818666643566556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,7168,0.04256355431344774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,65536,0.33698399861653644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,6144,0.053512000375323825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,6144,0.038776887787712946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,5120,0.033598221010631986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,5120,0.04461155666245354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,65536,0.4645582305060492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,4096,0.02718577782313029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,16384,0.1332471105787489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,4096,0.03571822245915731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,12288,0.10809244712193807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,3584,0.024654222859276667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,3584,0.03313866588804457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,10240,0.09441510836283366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,3072,0.022088888618681166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,3072,0.030349334081013996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,8192,0.0809840030140347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,2560,0.027494221925735474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,2560,0.01900800069173177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,2048,0.015421332584487068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,2048,0.025060445070266724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,7168,0.07496622535917494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,1536,0.011920889218648275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,1536,0.022668444448047217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,6144,0.06941155592600505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,1024,0.009299555586444007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,1024,0.020908445119857788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,5120,0.06331555710898505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,768,0.00811644477976693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,768,0.019855999284320407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,4096,0.053596443600124784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,512,0.006710222197903528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,512,0.01717511150572035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,3584,0.05092799994680616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,256,0.005010666532648934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,256,0.016195555528004963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,3072,0.046599109967549644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,128,0.004277333203289243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,128,0.015568888849682279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,2560,0.044166223870383374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,64,0.0038693332009845306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,64,0.015408888459205627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,32,0.0041235557032956015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,32,0.01566488875283135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,2048,0.040463109811147056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,1536,0.03837600019243028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,1024,0.03651466634538438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,768,0.03645066751374139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,16384,0.09700355264875625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,16384,0.07780000236299303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,65536,0.36824266115824383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,12288,0.07851288715998332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,12288,0.06249244345559014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,65536,0.30802665816413033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,10240,0.054246223635143705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,512,0.03171111146608988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,10240,0.0744675530327691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,256,0.031025777260462444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,8192,0.057124442524380155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,8192,0.04192444350984362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,128,0.029164443413416546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,7168,0.05266666743490431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,7168,0.03842133283615112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,6144,0.047974222236209445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,6144,0.03436266713672214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,5120,0.030107554462220933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,5120,0.04267288910018074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,65536,0.44850577248467344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,4096,0.02397955622937944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,4096,0.03411644366052415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,16384,0.13291466236114502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,3584,0.02195022172398037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,3584,0.03195644418398539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,12288,0.10878755648930867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,3072,0.019516444868511625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,3072,0.029566221767001685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,10240,0.09695911407470703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,2560,0.0163128889269299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,2560,0.027428444888856676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,8192,0.0825057758225335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,2048,0.01311377767059538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,2048,0.024681778417693243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,7168,0.07667111025916205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,1536,0.010747555229398938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,1536,0.022664000590642292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,6144,0.07023466957939996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,1024,0.008447999755541483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,1024,0.020058666666348774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,5120,0.06391377581490411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,768,0.007373332977294922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,768,0.018135999639829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,3584,0.05076266659630669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,512,0.00628977765639623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,512,0.0173688895172543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,4096,0.054126222928365074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,256,0.0047306666771570844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,256,0.01594577729701996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,3072,0.04593244526121351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,128,0.004085333396991094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,768,0.034550223085615374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,2560,0.04421155651410421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,128,0.015541333291265698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,64,0.0037297776175869834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,32,0.003984000119898054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,64,0.01533511115445031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,32,0.015580443872345818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,1536,0.03790311018625895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,2048,0.03988800115055508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,1024,0.035699556271235146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,65536,0.3606239954630534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,16384,0.09507022301355998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,16384,0.06689600149790446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,12288,0.0754986670282152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,12288,0.05500533183415731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,65536,0.2831946743859185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,10240,0.05080088973045349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,10240,0.06720533635881212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,8192,0.03767911261982388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,8192,0.05576088693406847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,256,0.03031555480427212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,7168,0.033512890338897705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,128,0.028391109572516546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,7168,0.05140799946255154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,6144,0.03065599997838338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,6144,0.0462355547481113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,512,0.03161511156294081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,5120,0.028608888387680054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,5120,0.04158755474620395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,65536,0.42754220962524414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,4096,0.02123733361562093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,4096,0.03308533297644721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,16384,0.1273804505666097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,3584,0.01940622263484531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,3584,0.031219555271996394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,12288,0.10485777590009902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,3072,0.017622222503026325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,3072,0.028472890456517536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,10240,0.09302310811148749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,2560,0.014548444085650973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,2560,0.026361778378486633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,8192,0.08375466532177395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,2048,0.012092444631788464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,2048,0.024297777149412367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,6144,0.06692355871200562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,1536,0.010075555907355415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,1536,0.022275555464956496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,7168,0.0738204452726576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,1024,0.007744889292452071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,1024,0.019168888529141743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,5120,0.06226933002471924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,768,0.007126222054163615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,768,0.01779022150569492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,3584,0.04861422379811605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,512,0.005523555394675996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,512,0.017041777571042378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,4096,0.052236444420284696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,256,0.004535111288229625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,256,0.016010666886965435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,3072,0.04439999990993076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,128,0.003953777667548922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,128,0.015384000208642749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,2560,0.04238222373856438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,64,0.0035626664757728577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,64,0.015427556302812366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,32,0.0038177776667806837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,768,0.03304355674319797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,32,0.01536711057027181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,1536,0.037309332026375666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,2048,0.03905866543451945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,65536,0.35315555996365017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,16384,0.09216622511545818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,16384,0.06319110923343234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,12288,0.05045066608322991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,65536,0.22099467118581137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,1024,0.035170667701297335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,12288,0.07355466816160414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,10240,0.0647617777188619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,10240,0.04442755712403191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,8192,0.03225955698225234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,8192,0.05401244428422716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,128,0.028102222416136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,256,0.030193779203626845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,7168,0.048998223410712353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,7168,0.03186310993300544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,6144,0.027626666757795546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,6144,0.04394133223427666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,512,0.030271109607484605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,5120,0.023746665981080797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,5120,0.038887110021379255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,16384,0.11241955227322048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,65536,0.3893795543246799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,4096,0.01851022243499756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,4096,0.032347556617524885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,3584,0.017105778058369953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,3584,0.030392888519499037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,12288,0.0928186641799079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,3072,0.014936000108718872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,3072,0.028394665982988145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,10240,0.08169333140055339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,2560,0.012633778154850006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,2560,0.026291555828518335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,8192,0.07093244128757052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,2048,0.011575111084514193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,2048,0.024108444650967915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,7168,0.06572088930341932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,1536,0.009303111169073317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,1536,0.022250667214393616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,6144,0.0596639977561103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,5120,0.055960887008243136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,1024,0.007497777541478475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,1024,0.018813333577579923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,768,0.006412444429265128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,768,0.017705778280893963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,3584,0.04614666766590542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,512,0.004961777892377642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,4096,0.047650665044784546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,512,0.016708445217874315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,256,0.004117333226733738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,256,0.015855999456511605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,3072,0.04325244492954678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,128,0.0038195554580953387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,128,0.015620445211728414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,2560,0.04045244389110141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,64,0.0034151110384199354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,64,0.015307555596033732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,32,0.003637333297067218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,32,0.015323556131786771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,2048,0.037717332442601524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,65536,0.3425288995107015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,1536,0.0363377763165368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,16384,0.057322667704688184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,65536,0.19178933567470977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,16384,0.08911199702156915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,768,0.03234044379658169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,12288,0.04725422130690681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,12288,0.07128800286187066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,1024,0.032925334241655134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,10240,0.06322133541107178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,10240,0.03647022114859687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,256,0.029740442832310993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,8192,0.029470221863852605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,8192,0.051557332277297974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,512,0.029578665892283123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,7168,0.047585778766208224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,7168,0.02718488872051239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,6144,0.02409155501259698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,128,0.028235554695129395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,6144,0.04271822174390157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,5120,0.020670221911536325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,5120,0.03762133253945245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,65536,0.393812444474962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,4096,0.017924444542990792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,16384,0.11115645037757026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,4096,0.03185333477126227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,3584,0.01536177761024899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,3584,0.03001333276430766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,12288,0.09072799815071954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,3072,0.013394667042626275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,3072,0.028090665737787884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,10240,0.08033955759472318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,2560,0.011742221812407175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,2560,0.026055999928050574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,8192,0.06961777475145128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,2048,0.010039110978444418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,2048,0.024043555061022442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,7168,0.06422577963935004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,1536,0.008629333641793992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,1536,0.02169688873820835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,6144,0.06047822369469536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,1024,0.006933333145247565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,1024,0.019096000327004325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,5120,0.05518488751517402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,768,0.005852444304360284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,768,0.017349332571029663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,4096,0.04792355497678121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,3584,0.045718222856521606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,512,0.004884444591071871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,512,0.016759999924235873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,256,0.0039057777159743835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,256,0.0158595558669832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,3072,0.043042666382259794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,128,0.003530666646030214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,128,0.015283554792404175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,2560,0.04072177741262648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,64,0.003206222214632564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,64,0.015119110544522604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,32,0.0033937777496046494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,32,0.015110222829712762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,2048,0.03794666793611314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,1536,0.0361697773138682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,1024,0.03382222188843621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,768,0.03261422117551168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,16384,0.08715733554628159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,16384,0.05102133419778612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,65536,0.34069156646728516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,12288,0.07039911217159696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,12288,0.042100446091757886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,65536,0.16621866491105822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,10240,0.061455110708872475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,10240,0.038570665650897555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,8192,0.028304000695546467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,8192,0.05085600084728665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,128,0.028103109863069322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,256,0.029653333955340918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,512,0.03014311194419861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,7168,0.04583644535806444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,7168,0.02531822191344367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,6144,0.04145688811937968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,6144,0.022537777821222942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,5120,0.019931554794311523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,5120,0.03730400072203742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,16384,0.11507288614908855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,4096,0.015286222100257874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,4096,0.03202133377393087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,65536,0.3902568817138672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,3584,0.013696889082590738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,3584,0.030285331938001845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,10240,0.08379822307162815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,12288,0.09508977995978461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,3072,0.012854222622182635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,3072,0.027895109521018133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,2560,0.01089866707722346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,2560,0.025992888543340895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,8192,0.07252622312969632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,2048,0.00943199959066179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,2048,0.023620444867346022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,6144,0.06041510899861654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,7168,0.06605333089828491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,1536,0.007925333248244392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,1536,0.020628444022602506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,1024,0.0063742221229606206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,1024,0.018565333551830716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,5120,0.05448711249563429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,768,0.005060444275538127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,768,0.017495999733606975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,4096,0.04740533232688904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,512,0.004335111214054955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,512,0.0166302224000295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,3584,0.04474044508404202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,256,0.0038808890514903595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,256,0.015990222493807476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,2560,0.0396071109506819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,128,0.0034533333447244433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,128,0.015311110350820752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,3072,0.041760888364579946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,64,0.0031831111345026228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,64,0.015051555302408008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,32,0.0033493333806594214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,32,0.014959999256663851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,2048,0.037568890386157565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,65536,0.14503555827670628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,65536,0.33710400263468426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,16384,0.0447324448161655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,768,0.03133066826396518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,1536,0.03478399912516276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,1024,0.03129511078198751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,16384,0.08538311057620579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,12288,0.0682328873210483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,12288,0.03684711125161912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,10240,0.06027999851438734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,10240,0.0329146683216095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,8192,0.023337778117921617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,8192,0.050160888168546885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,256,0.029623998536003962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,128,0.02788177794880337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,7168,0.04496888981925117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,7168,0.02055288851261139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,6144,0.0180711117055681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,6144,0.03993777765168084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,512,0.02938755684428745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,5120,0.017755554782019723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,5120,0.036582221587498985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,16384,0.10355999734666611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,65536,0.3651306629180908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,4096,0.01404266721672482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,4096,0.03137866655985514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,12288,0.08480088578330146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,3584,0.01275466630856196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,3584,0.029651555750105116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,3072,0.012030222349696688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,3072,0.02758933272626665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,10240,0.07518311341603597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,2560,0.010092444717884064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,2560,0.025807999902301367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,8192,0.06557955344518025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,2048,0.008882666627566019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,2048,0.023620444867346022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,7168,0.0603297750155131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,1536,0.0074622225430276655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,1536,0.020311999652120803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,6144,0.056258665190802686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,1024,0.0057955553962124725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,1024,0.01816444430086348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,5120,0.05128088924619886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,768,0.004763555609517627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,768,0.017078222499953378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,3584,0.043430222405327686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,4096,0.045180443260404796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,512,0.004179555508825514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,512,0.016484444340070088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,256,0.0037164443896876443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,3072,0.04096355703141954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,256,0.01564711166752709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,128,0.003400000019205941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,128,0.014904000692897372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,2560,0.03903822104136149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,64,0.0031475555151700974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,64,0.015025777949227227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,2048,0.03663911090956794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,32,0.0033404444240861465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,32,0.015209777487648858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,65536,0.11691289477878147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,65536,0.33711555269029403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,1536,0.03439733386039734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,16384,0.03763822383350796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,16384,0.0836355553732978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,768,0.03147911032040914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,1024,0.0315048893292745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,12288,0.0668720006942749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,12288,0.03506933318244086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,10240,0.05951644314659966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,10240,0.031142221556769475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,8192,0.019070222973823547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,8192,0.04934044347869026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,128,0.028202666176689997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,256,0.02983377708329095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,7168,0.019567999574873183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,7168,0.043987555636299976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,512,0.029117332564459905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,6144,0.018632888793945312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,6144,0.040638221634758845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,5120,0.01589777734544542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,5120,0.03639199998643663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,12288,0.08109066883722942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,16384,0.09831288788053726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,4096,0.01075111081202825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,4096,0.031480888525644936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,3584,0.009964444571071202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,3584,0.029616889026429918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,10240,0.07196088631947835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,3072,0.009899555808968013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,65536,0.33780709902445477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,3072,0.027489778068330552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,2560,0.008638222184446123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,2560,0.02474844455718994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,8192,0.06173244449827406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,2048,0.007282666862010956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,2048,0.02196888956758711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,7168,0.05745422177844577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,6144,0.052645332283443876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,1536,0.006524444454246097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,1536,0.019966221517986722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,1024,0.005264889034960005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,1024,0.018106665876176622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,5120,0.048835555712381996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,768,0.004496888981925117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,768,0.01717866626050737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,4096,0.04353777898682488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,512,0.004078222231732475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,3584,0.0421653323703342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,512,0.016524444023768108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,256,0.003632889025741153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,256,0.01568711135122511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,3072,0.039672000540627375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,128,0.003362666608558761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,128,0.0151235560576121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,2560,0.037808001041412354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,64,0.0031715554909573663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,64,0.01494488947921329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,32,0.003354666754603386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,32,0.014895111322402954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,2048,0.03555999861823188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,65536,0.3367351161109076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,65536,0.09209599759843613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,16384,0.03119644522666931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,16384,0.0841404464509752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,1536,0.033424887392256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,1024,0.03057866626315647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,12288,0.02678311036692725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,12288,0.06681066751480103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,768,0.030491554074817236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,10240,0.024280889166726008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,10240,0.057723555299970836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,8192,0.017273777061038546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,8192,0.04858666658401489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,256,0.02931555443339878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,512,0.029000888268152874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,7168,0.015473778049151102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,128,0.02760444415940179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,7168,0.044533332188924156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,6144,0.013798221945762634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,6144,0.040463109811147056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,5120,0.01460533340771993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,5120,0.03640800052218967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,65536,0.33292799525790745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,4096,0.01092266705301073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,4096,0.03133866522047255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,16384,0.09526222281985813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,3584,0.010236444572607676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,3584,0.029531555043326482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,12288,0.07926755481296115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,3072,0.009836444424258338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,3072,0.02742133372359806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,10240,0.07045866383446588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,2560,0.008190222084522247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,8192,0.06066311068005032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,2560,0.02421155571937561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,2048,0.007012444237867991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,2048,0.02164711058139801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,6144,0.05118844575352139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,7168,0.055917335881127246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,1536,0.006026666611433029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,1536,0.019875556230545044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,1024,0.005162666655249066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,1024,0.018044445249769423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,5120,0.04842755529615614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,768,0.004345777961942884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,768,0.01702044407526652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,4096,0.04342222213745117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,512,0.0038604442444112566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,512,0.016141333513789706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,3584,0.041526221566730075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,256,0.003571555432346132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,256,0.015581332974963717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,3072,0.039072000318103366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,128,0.003232888877391815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,128,0.01497155593501197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,2560,0.035544888840781316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,64,0.0030586665703190696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,64,0.014814222852389017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,32,0.0031679999083280563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,32,0.014829332629839579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,65536,0.06792977783415052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,2048,0.034509334299299456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,65536,0.3350035614437527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,16384,0.025783111651738484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,16384,0.08446133136749268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,1536,0.03195200032658047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,12288,0.0207022229830424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,1024,0.030414223670959473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,12288,0.06676889128155179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,10240,0.01757066614098019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,768,0.030588444736268785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,10240,0.05886489152908325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,8192,0.014031999640994601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,8192,0.04845866560935974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,256,0.02918133470747206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,512,0.028676443629794653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,7168,0.013222222526868185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,7168,0.0439626673857371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,6144,0.014344889256689282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,6144,0.04012088974316915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,128,0.027429333991474573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,5120,0.01293777757220798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,5120,0.036166220903396606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,65536,0.3225795427958171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,4096,0.011391110718250275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,4096,0.031027555465698242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,16384,0.09275377458996242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,3584,0.00979288915793101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,3584,0.027951998843087092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,12288,0.07687911060121325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,3072,0.009161777794361115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,3072,0.02625511089960734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,8192,0.058871997727288135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,2560,0.008023999631404877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,10240,0.0682471129629347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,2560,0.023876445160971746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,2048,0.0069617778062820435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,2048,0.021740444832377966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,7168,0.054021331999037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,1536,0.0058453331391016645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,6144,0.0503022207154168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,1536,0.019713777634832595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,1024,0.004950222041871813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,1024,0.017960889471901786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,5120,0.04756888747215271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,768,0.004327110946178436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,768,0.01718488832314809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,4096,0.042612443367640175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,512,0.0038657776183552216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,512,0.01629066632853614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,3584,0.04049866729312473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,256,0.003370666669474708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,256,0.015396444333924187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,3072,0.03621777892112732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,128,0.003097777772280905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,128,0.014785778191354541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,2560,0.0350471105840471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,2048,0.03245866629812453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,64,0.003027555636233754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,32,0.0030986666679382324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,64,0.015016888578732809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,32,0.014859555496109856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,65536,0.3378426763746474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,65536,0.053949332899517484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,16384,0.01905866629547543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,16384,0.08437155352698432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,1536,0.03131644593344794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,1024,0.03010666701528761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,12288,0.0155395550860299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,12288,0.06566933128568861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,10240,0.014910222755538093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,768,0.03048622277047899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,10240,0.05718133184644911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,8192,0.013234666652149625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,8192,0.0476746658484141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,512,0.02896889050801595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,256,0.02910311023394267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,7168,0.012542222109105853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,7168,0.04398399922582838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,6144,0.012270222107569376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,6144,0.03991288940111796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,128,0.02734755476315816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,5120,0.011036444041464063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,5120,0.035841776265038386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,16384,0.09220266342163086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,4096,0.00943022221326828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,65536,0.32077956199645996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,4096,0.030528889762030706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,3584,0.008544888761308458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,3584,0.028524445162879095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,12288,0.07621422078874376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,3072,0.007849777738253275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,3072,0.0255039996571011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,10240,0.06846933232413398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,2560,0.007376888559924231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,2560,0.023696889479955036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,8192,0.058061334821912974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,2048,0.00600888869828648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,2048,0.02159822152720557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,7168,0.05358933409055074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,1536,0.005391110976537068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,1536,0.01971733404530419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,6144,0.050145778391096324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,1024,0.00452977791428566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,1024,0.018007111218240526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,5120,0.04749600092569987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,768,0.0042071110672420925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,768,0.017088888419999015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,4096,0.041972445117102734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,512,0.0037644443412621817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,512,0.016355555918481614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,3584,0.040234665075937905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,256,0.0033982222278912864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,256,0.015669332610236276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,3072,0.037074665228525795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,128,0.00309599998096625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,128,0.014875556031862894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,2560,0.03538844320509169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,64,0.0029680000411139596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,64,0.014897776974572076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,2048,0.03226133187611898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,32,0.0030168888883458245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,32,0.014827556080288358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,65536,0.04936088787184822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,65536,0.33491023381551105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,16384,0.014134221606784396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,16384,0.08453333377838135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,1536,0.031160000297758315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,12288,0.01203644441233741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,12288,0.06590488884184095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,1024,0.030058665408028498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,10240,0.01146311147345437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,10240,0.05693866809209188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,768,0.030438221163219873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,8192,0.009890666438473595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,8192,0.04774666825930277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,512,0.02865155537923177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,7168,0.011614222493436603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,256,0.02939200070169237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,7168,0.04387199878692627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,6144,0.009734222458468543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,6144,0.03959200117323134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,128,0.027279999521043565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,5120,0.010103111465771994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,5120,0.03527822097142538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,16384,0.09113688601387872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,4096,0.008801778157552084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,4096,0.02958933181232876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,65536,0.3177493413289388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,3584,0.007707555260923173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,3584,0.027906666199366253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,12288,0.07502488957511054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,3072,0.007369777394665613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,3072,0.025263999899228413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,10240,0.06718844175338745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,2560,0.006516444600290722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,2560,0.023372444841596816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,8192,0.05691111087799072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,2048,0.0058400001790788435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,2048,0.02163822286658817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,7168,0.053114665879143604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,1536,0.005105777747101254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,6144,0.0496417780717214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,1536,0.019761777586407132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,1024,0.004376888689067629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,1024,0.017835555805100333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,5120,0.04742133286264208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,768,0.004089777668317159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,768,0.016758221719000075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,4096,0.03965955641534593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,512,0.003699555579158995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,3584,0.03930933276812235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,512,0.016152888536453247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,256,0.003335111050142182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,256,0.015392889579137167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,3072,0.03509155578083462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,2560,0.03454400102297465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,128,0.003083555648724238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,128,0.015021334091822306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,2048,0.03201866812176175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,64,0.002967111145456632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,64,0.014797333214018079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,32,0.0029680000411139596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,32,0.014729777971903482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,65536,0.3369742234547933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,65536,0.038331554995642766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,16384,0.011877333124478659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,16384,0.08363378047943115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,1024,0.030208888981077407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,1536,0.03122577733463711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,12288,0.01036444471942054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,12288,0.06688800123002794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,10240,0.010398222340477837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,10240,0.058203558127085366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,768,0.030335999197430078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,8192,0.008592888712882996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,8192,0.047112001313103564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,512,0.028911110427644517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,7168,0.008046222229798635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,7168,0.04316622349951002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,256,0.028974221812354192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,6144,0.009755555954244401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,6144,0.03863911165131463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,128,0.027496889233589172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,5120,0.008823999928103553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,5120,0.03453866640726725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,16384,0.09087555276023017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,65536,0.31622934341430664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,4096,0.007879111501905654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,4096,0.029969778325822618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,3584,0.00755466686354743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,3584,0.027923554182052612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,12288,0.07471466726726957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,3072,0.007165333463086023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,3072,0.025377778543366328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,10240,0.06659822331534491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,2560,0.006430222342411677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,2560,0.023527999718983967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,8192,0.05647999710506863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,2048,0.005656888915432825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,2048,0.021365332934591506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,7168,0.05270488725768196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,1536,0.004977777600288391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,1536,0.019672888848516677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,6144,0.0483857790629069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,1024,0.004282666577233208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,1024,0.01774666706720988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,5120,0.04652888907326592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,768,0.003989333493842019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,768,0.016910221841600206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,4096,0.03920000129275852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,512,0.003589333345492681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,3584,0.03840000099605984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,512,0.015986666083335876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,256,0.0032791110376516977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,256,0.015216888652907478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,3072,0.03503377901183234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,128,0.0030622221529483795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,128,0.014906666345066495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,2560,0.033955554167429604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,64,0.002951111023624738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,64,0.014674666855070325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,32,0.003042666655447748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,32,0.01460000044769711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,2048,0.03167377909024557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,65536,0.02570488883389367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,65536,0.33700177404615617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,16384,0.010253333383136326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,16384,0.08419466680950588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,1536,0.03092799915207757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,12288,0.009100444614887238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,12288,0.065610667069753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,1024,0.029725333054860432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,10240,0.008457777400811514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,10240,0.05670666694641113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,768,0.030233777231640283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,8192,0.007479111353556315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,8192,0.04675466815630595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,512,0.028380443652470905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,7168,0.007229333122571309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,256,0.028824888997607764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,7168,0.04274310999446445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,6144,0.00699733363257514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,6144,0.037556442949506975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,128,0.027297778262032404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,5120,0.007393777370452881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,5120,0.03358666764365302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,65536,0.3174613316853841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,4096,0.007056000332037608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,4096,0.029260443316565618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,16384,0.09080977572335137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,3584,0.007332444190979004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,3584,0.027064000566800434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,12288,0.07493955559200711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,3072,0.007044444481531779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,3072,0.02529777752028571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,10240,0.06565244330300225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,2560,0.006215999937719769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,2560,0.023588443795839947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,8192,0.056779556804233126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,2048,0.005577777822812398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,2048,0.021536888347731695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,7168,0.05225066675080193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,1536,0.004836444639497333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,1536,0.019476445184813607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,6144,0.04868177904023064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,1024,0.0042071110672420925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,1024,0.017690666847758822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,5120,0.04424800144301521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,768,0.0038720000949170855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,768,0.01678400072786543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,3584,0.037243554989496865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,512,0.0034968890249729156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,4096,0.0388426681359609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,512,0.015835555063353646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,256,0.003181333343187968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,256,0.015112888481881885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,3072,0.03475822342766656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,128,0.002979555477698644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,128,0.014478221535682678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,2560,0.03385333220163981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,64,0.0028311111446883944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,64,0.014557333456145393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,32,0.00282222218811512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,32,0.014463999205165438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,65536,0.022738666998015508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,65536,0.3363200028737386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,16384,0.007730666961934831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,16384,0.08321511083179049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,12288,0.007162666983074612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,12288,0.06539377901289198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,10240,0.007244444555706448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,10240,0.0565626687473721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,8192,0.007092444433106317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,8192,0.04637955625851949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,7168,0.007060444189442529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,7168,0.042688889635933765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,6144,0.006827555596828461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,6144,0.03766311208407084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,5120,0.00721688899728987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,5120,0.03328444560368856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,4096,0.006779555645253923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,4096,0.02895911203490363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,3584,0.007262222468852997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,3584,0.02709688908523983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,3072,0.006888888776302338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,3072,0.02516088883082072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,2560,0.006142222219043308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,2560,0.02328355610370636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,2048,0.005474666754404704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,2048,0.021431111627154883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,1536,0.00481244424978892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,1536,0.019468444916937087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,1024,0.004134222037262387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,1024,0.017606221967273288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,768,0.003802666647566689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,768,0.01664355562792884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,2048,0.031926221317715116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,512,0.0034604445099830627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,256,0.003104888937539525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,512,0.015974221958054435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,128,0.00299822228650252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,256,0.015121777852376303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,64,0.0027919999427265594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,128,0.01475911173555586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,32,0.00286666676402092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,64,0.014519110321998596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,65536,0.015575110912322998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,32,0.01462399959564209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,16384,0.007088888850477006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,1536,0.030680888228946265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,12288,0.0069208890199661255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,16384,0.08390221993128459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,10240,0.00721599989467197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,65536,0.3362897766960992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,12288,0.06501599815156725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,8192,0.0070773329999711775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,10240,0.05645510885450575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,7168,0.006918222126033571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,8192,0.04552177919281853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,6144,0.006717333363162146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,7168,0.04224088788032532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,5120,0.007101332975758447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,6144,0.03675999906327989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,1024,0.02950577934583028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,4096,0.006681777950790193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,5120,0.03334577878316244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,3584,0.00720088928937912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,4096,0.02888266576661004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,3072,0.00685066646999783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,3584,0.027150221996837195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,2560,0.006155555446942647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,3072,0.025314667158656653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,2048,0.005487999982304043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,2560,0.02331022255950504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,1536,0.004872888740566042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,2048,0.02126755482620663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,1024,0.0041831110914548235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,1536,0.019463999403847587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,768,0.0038373333712418876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,1024,0.01761511133776771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,512,0.0035048888789282907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,768,0.016661332713233102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,256,0.0031831111345026228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,512,0.015820443630218506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,128,0.0029599999801980127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,256,0.015122666954994202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,64,0.002806222273243798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,128,0.01460533340771993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,32,0.0028133332315418455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,64,0.01442488862408532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,32,0.014336888988812765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,768,0.0303146673573388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,512,0.028255999088287354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,256,0.028860443168216284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,128,0.02716355522473653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,16384,0.5529511239793565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,12288,0.4181982146369086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,10240,0.35989422268337673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,12288,0.44615734948052305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,16384,0.5614257918463813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,10240,0.5013875431484646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,8192,0.2887609004974365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,10240,0.3847457832760281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,8192,0.31319644716050893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,7168,0.25858044624328613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,12288,0.5925173229641384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,6144,0.22574755880567762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,7168,0.3507866594526503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,8192,0.3983200126224094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,7168,0.27666844262017143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,6144,0.24331466356913248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,16384,0.776634693145752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,6144,0.3074337906307644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,5120,0.18457155757480195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,5120,0.19758221838209364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,4096,0.1497911082373725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,5120,0.2589297824435764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,3584,0.13641511069403753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,4096,0.16810222466786703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,3584,0.1488800048828125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,2560,0.09991111358006795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,4096,0.21046222580803764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,3072,0.11585599846310085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,3584,0.21236799822913277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,3072,0.13670221964518228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,2560,0.11708711253272162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,2560,0.13974311616685656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,3072,0.1640017827351888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,2048,0.08645333184136285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,1536,0.06677955389022827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,1024,0.05470577875773112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,2048,0.10338311062918769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,1024,0.0656577746073405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,1536,0.08355110883712769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,768,0.04363911019431221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,1024,0.06902666886647542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,2048,0.11543199751112197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,1536,0.09203110800849067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,768,0.06099200248718262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,512,0.03802400165134006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,512,0.05507911245028178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,128,0.018606222338146634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,256,0.03048444456524319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,512,0.04101244608561198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,128,0.028041776683595445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,768,0.053426666392220386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,128,0.04066933194796244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,64,0.014009777042600842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,256,0.04714399907324049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,32,0.0144195556640625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,64,0.029431111282772485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,32,0.03427199853791131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,256,0.02492711113558875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,12288,0.1304533349143134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,16384,0.1666568915049235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,16384,0.1918648878733317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,12288,0.14972533120049372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,10240,0.1091831127802531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,12288,0.15299910969204372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,10240,0.12615466117858887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,10240,0.1331982215245565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,16384,0.19525067011515299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,8192,0.09103555811776055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,8192,0.1130782233344184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,7168,0.08232355780071683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,8192,0.10117333465152317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,7168,0.10343733098771836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,7168,0.10321244266298081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,6144,0.07337333096398248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,6144,0.09304977787865533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,6144,0.07892622550328572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,5120,0.061956445376078285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,4096,0.052481777138180204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,5120,0.06682666805055407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,65536,0.6343590948316786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,5120,0.08009777466456096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,4096,0.054901334974500865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,3584,0.04783111148410373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,4096,0.06817599799897936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,3584,0.06474400228924222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,3584,0.04866844415664673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,3072,0.043383111556371055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,3072,0.04329244295756022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,2560,0.038535998927222356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,3072,0.06020800272623698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,2048,0.034204443295796715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,2560,0.05618488788604736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,2560,0.03717689050568475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,2048,0.03103733393881056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,1536,0.030116445488399927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,1024,0.02381333377626207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,1536,0.04597599969969856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,2048,0.050408889849980675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,768,0.013722666435771518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,1024,0.04169955518510607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,768,0.022207111120224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,512,0.010237333675225576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,768,0.03877866599294875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,1024,0.01759733259677887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,512,0.0206995556751887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,256,0.00756177802880605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,128,0.005578666511509154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,512,0.03750666644838121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,256,0.01796800063716041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,1536,0.02494222256872389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,128,0.016520000166363187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,64,0.004947555561860402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,256,0.03434755404790243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,128,0.029783109823862713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,32,0.0053013331360287136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,64,0.017017778423097398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,32,0.017782222893502977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,65536,0.6937564214070638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,16384,0.14649689197540283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,16384,0.15105066034528944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,16384,0.1468284395005968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,12288,0.12243555651770698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,12288,0.11707822481791179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,12288,0.12057956059773763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,10240,0.09794044494628906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,10240,0.09549244244893391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,10240,0.10759911272260879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,8192,0.08075822061962552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,8192,0.09284800291061401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,7168,0.07314666774537829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,8192,0.0770675539970398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,7168,0.06823199987411499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,7168,0.08659199873606364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,65536,0.5518915388319228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,6144,0.06594044632381864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,65536,0.5512160195244683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,6144,0.07745866643057929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,6144,0.06052622530195448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,5120,0.0509786672062344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,5120,0.05679555733998617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,4096,0.04749244451522827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,4096,0.05932799975077311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,5120,0.06905688842137654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,3584,0.04317155480384827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,4096,0.047585778766208224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,3584,0.03774044579929776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,3072,0.03940355446603563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,2560,0.035064889325035944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,3584,0.05706222189797295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,3072,0.05387022097905477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,3072,0.033364444971084595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,2048,0.03122044603029887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,2560,0.05048533280690511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,2048,0.04471110966470507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,1536,0.026657778355810378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,1536,0.04170666800604927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,1024,0.012997333374288348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,1024,0.023027555810080633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,2048,0.023893333143658105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,2560,0.02888710962401496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,768,0.010760889285140567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,1536,0.01901333365175459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,768,0.022061333060264587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,1024,0.03926755653487311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,512,0.008586666650242275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,512,0.02033955521053738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,256,0.006315555423498154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,768,0.03647999962170919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,65536,0.7505262162950305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,256,0.017866666118303936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,128,0.004654222064548069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,512,0.036299556493759155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,256,0.031679110394583807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,128,0.01664888858795166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,64,0.0042053332759274375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,32,0.017401778035693698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,128,0.028782222006056044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,32,0.004574222283230888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,64,0.016901334126790363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,16384,0.13971466488308376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,16384,0.14227999581231013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,12288,0.10010577572716607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,65536,0.6084168752034506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,16384,0.12949600484636095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,12288,0.11022844579484727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,10240,0.09261155790752834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,10240,0.10127377510070801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,12288,0.1159084505505032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,8192,0.06681955522961087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,10240,0.08982221947775947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,8192,0.07732799980375502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,8192,0.08720266819000244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,7168,0.06953955358929105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,7168,0.08614133463965522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,6144,0.06289066871007283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,7168,0.060191108120812305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,6144,0.055128888951407544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,6144,0.07351111041174994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,5120,0.05309244328074985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,5120,0.044844445255067616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,4096,0.045679999722374805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,5120,0.0661075578795539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,4096,0.057088891665140785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,4096,0.036791112687852644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,3072,0.03797244363360935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,3584,0.04169333312246534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,65536,0.5448675685458714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,3072,0.051375110944112144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,3584,0.054851555162005954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,2560,0.03436177637841966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,3072,0.029342220889197454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,3584,0.03302222159173753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,2560,0.047784000635147095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,2048,0.03014844324853685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,1536,0.01676977839734819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,2048,0.042824887567096286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,1536,0.02604533400800493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,1024,0.011626666618718041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,2560,0.025268443756633337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,1536,0.040024889840020075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,1024,0.022608000371191237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,768,0.009879111415810054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,1024,0.038246221012539335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,768,0.021351110604074266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,2048,0.021347555849287245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,512,0.007833778030342525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,768,0.03606933355331421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,512,0.018817777434984844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,256,0.00565155554148886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,512,0.034670223792394005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,256,0.017475555340449016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,128,0.004592000196377437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,256,0.030270222160551283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,128,0.016427555018001132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,64,0.00407644444041782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,128,0.028414222929212783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,32,0.004524444540341695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,64,0.016717332932684157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,32,0.016996444927321542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,65536,0.5145911110772027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,16384,0.102856887711419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,12288,0.08032088809543185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,12288,0.08254755867852105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,16384,0.10212444596820408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,10240,0.0728986660639445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,10240,0.06875022252400716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,65536,0.3869200017717149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,8192,0.05927111042870415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,65536,0.4993875291612413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,8192,0.0537484453784095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,7168,0.05450400047832065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,7168,0.048948443598217435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,6144,0.04339822133382162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,6144,0.05166755451096428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,65536,0.43284977806939023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,5120,0.036687112516827054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,5120,0.04398399922582838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,16384,0.12654844919840494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,4096,0.028976887464523315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,4096,0.03727644350793626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,12288,0.10462222496668498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,3584,0.02616888946957058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,3584,0.03421866562631395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,10240,0.09322933355967204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,3072,0.023489778240521748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,3072,0.03142933381928338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,65536,0.43903644879659015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,2560,0.02258311046494378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,8192,0.08099022176530626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,2560,0.0277804434299469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,2048,0.016357334123717416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,2048,0.025182222326596577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,7168,0.07691644297705756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,1536,0.012723555167516073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,1536,0.023038221730126276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,6144,0.06880444288253784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,1024,0.009586667021115622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,1024,0.021125333176718816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,5120,0.06162133481767443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,768,0.008221333225568136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,768,0.019110222657521565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,3584,0.05073777834574381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,4096,0.053256889184316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,512,0.006903111106819577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,512,0.01759288873937395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,256,0.004537777768241035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,256,0.016200888488027785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,2560,0.04451466600100199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,128,0.004045333299371931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,128,0.015517334143320719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,1536,0.03836444351408217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,64,0.003671110918124517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,64,0.015466666883892484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,32,0.004013333469629288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,32,0.01571555601225959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,3072,0.047707554366853505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,2048,0.03988800115055508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,768,0.03503911031617059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,1024,0.03702133231692844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,16384,0.09772444433636135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,65536,0.36492000685797793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,16384,0.09364000293943618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,65536,0.4787591298421224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,12288,0.07915733257929485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,16384,0.13230933083428278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,12288,0.06922488742404513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,12288,0.10555999808841282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,10240,0.06751200225618151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,10240,0.05801066425111559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,8192,0.0563946697447035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,10240,0.09019288751814102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,8192,0.04690666662322151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,8192,0.07697244485219319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,7168,0.05246222350332472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,7168,0.07277955611546834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,7168,0.04235466652446323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,256,0.0298977792263031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,6144,0.047909332646263965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,6144,0.03738044367896186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,6144,0.06628088818656074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,5120,0.040928890307744346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,4096,0.035869333479139544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,65536,0.36447466744316953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,4096,0.05128888951407539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,5120,0.05876533190409342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,4096,0.02624799973434872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,5120,0.03208177619510227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,3584,0.03355111016167535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,3584,0.04838933216200935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,3072,0.030132446024152968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,3072,0.04670666654904684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,128,0.028128001425001357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,3584,0.026124444272783067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,2560,0.02825777729352315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,2048,0.015138667490747241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,2560,0.0438435541258918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,2048,0.025008888708220586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,1536,0.011743110915025076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,1536,0.022887999812761944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,2048,0.03950400153795878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,3072,0.02125066684352027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,2560,0.018202667435010273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,1024,0.00909511082702213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,1536,0.03841777642567953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,768,0.007847111258241866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,1024,0.020418667131000094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,768,0.01921866668595208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,1024,0.03714488943417867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,512,0.006732444382376141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,768,0.03477777706252204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,512,0.01719999975628323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,256,0.004754666652944353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,512,0.03183644347720676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,256,0.016336000627941556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,128,0.00424799985355801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,256,0.0302746660179562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,128,0.028207110034094915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,128,0.01534577707449595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,64,0.003908444609906939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,32,0.0042151109211974675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,64,0.015364444918102689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,32,0.015594666202863058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,512,0.03238666719860501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,16384,0.0935608877076043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,12288,0.06000533368852404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,12288,0.0753564437230428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,16384,0.08769422107272679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,10240,0.06454222069846259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,10240,0.05769244167539808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,8192,0.054049776660071484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,8192,0.04096444447835287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,65536,0.35028354326883954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,7168,0.05008888906902737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,7168,0.03698755635155572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,6144,0.03281599945492215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,6144,0.046005331807666354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,65536,0.31626934475368923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,5120,0.028179556131362915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,5120,0.03977510995335049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,16384,0.1121857828564114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,12288,0.09161600139405991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,10240,0.08085155487060547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,4096,0.03387466735310025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,4096,0.02330933345688714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,3584,0.0314631097846561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,3584,0.021048888564109802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,8192,0.0708106689982944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,3072,0.029393778906928167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,3072,0.019017777509159513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,2560,0.015882665912310284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,2560,0.026872000760502283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,7168,0.06763199965159099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,2048,0.013640888863139682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,2048,0.02416977783044179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,65536,0.3763608932495117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,1536,0.010417777631017896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,1536,0.02219911085234748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,6144,0.061231997278001576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,1024,0.008219555848174626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,1024,0.0203448881705602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,5120,0.05507733424504598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,768,0.007168889045715332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,768,0.018195556269751657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,4096,0.047820445564058095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,512,0.0058337777025169795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,512,0.017239999439981248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,3584,0.04629244407018026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,256,0.004302222281694412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,256,0.015995555453830294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,2560,0.042615112331178456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,3072,0.04489955637190077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,128,0.0038577777643998465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,128,0.015231110983424716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,64,0.0035217776894569397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,64,0.015300444430775113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,32,0.003779555360476176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,32,0.015470221638679504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,2048,0.03888444436921014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,1536,0.03753777676158481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,1024,0.037265777587890625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,65536,0.3375386661953396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,16384,0.07395733065075345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,16384,0.0889528857337104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,768,0.035212443934546575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,12288,0.0698524448606703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,12288,0.056332442495557994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,65536,0.2558657858106825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,10240,0.06292000081804064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,10240,0.044144888718922935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,8192,0.039121776819229126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,8192,0.05260711246066623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,256,0.02943555514017741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,512,0.03197333216667175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,7168,0.04741422335306803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,7168,0.03198755449718899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,128,0.027887112564510767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,6144,0.04212355613708496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,6144,0.030356443590588037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,5120,0.024528000089857314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,5120,0.036158220635520086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,16384,0.12300622463226318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,65536,0.4453760253058539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,4096,0.032855110036002264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,4096,0.02037333283159468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,3584,0.018426666657129925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,3584,0.030927111705144245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,12288,0.09891377554999457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,3072,0.016159110599093966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,3072,0.02740266587999132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,10240,0.08614222208658855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,2560,0.013501333693663279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,2560,0.025548444853888616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,8192,0.07428177860048082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,2048,0.012568000290128918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,2048,0.023656888140572443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,7168,0.06893599695629544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,1536,0.009398221969604492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,1536,0.021742223037613764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,6144,0.06333333253860474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,1024,0.00815555536084705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,1024,0.0192266669538286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,5120,0.05609333515167236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,768,0.006983111302057902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,768,0.017883555756674874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,4096,0.048957334624396436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,512,0.005006222261322869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,512,0.01721955504682329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,3584,0.0458568897512224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,256,0.004320888883537716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,256,0.016158221496476066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,3072,0.04474400149451362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,128,0.0037208890749348533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,128,0.015391111373901367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,2560,0.041131556034088135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,64,0.003568888952334722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,64,0.015272888872358533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,32,0.0036311112344264984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,32,0.015192889504962497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,2048,0.03808266586727566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,1536,0.0365608897474077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,1024,0.035114665826161705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,768,0.0310204459561242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,65536,0.324727111392551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,16384,0.08536355362998115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,16384,0.05783910883797539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,65536,0.21067467000749376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,12288,0.06886933247248332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,12288,0.0439280006620619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,10240,0.058527999454074435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,10240,0.05085866649945577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,8192,0.029472000069088403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,8192,0.050360000795788236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,512,0.03140711122088962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,256,0.02986488739649455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,7168,0.02960977620548672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,7168,0.04564266734653049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,128,0.028303109937243994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,6144,0.03997155692842271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,6144,0.025013332565625507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,5120,0.022103110949198406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,5120,0.035191112094455294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,16384,0.10747288995318943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,4096,0.0165928883685006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,4096,0.031209776798884075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,12288,0.08919999996821086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,3584,0.014713777436150445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,3584,0.028986665937635634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,65536,0.3418568770090739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,3072,0.014157333307796054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,3072,0.027202667461501226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,10240,0.07858577701780531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,2560,0.01162755572133594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,2560,0.02502044373088413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,8192,0.0683271090189616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,2048,0.00981600003110038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,2048,0.023386667172114056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,7168,0.06503911150826348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,6144,0.058690667152404785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,1536,0.008575999902354347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,1536,0.020391111572583515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,1024,0.006807111203670502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,1024,0.01891644464598762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,5120,0.05194133188989428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,768,0.0058417779703934984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,768,0.018000000052981906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,3584,0.04334222277005514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,512,0.00481244424978892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,4096,0.045420444673962064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,512,0.01705511079894172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,256,0.004044444610675176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,256,0.015993777248594496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,3072,0.04179999894566006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,128,0.0035955554081334006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,128,0.015357333752844067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,2560,0.03968800107638041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,64,0.003286222202910317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,64,0.015091554986106025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,32,0.003499555504984326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,32,0.014948444234000312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,2048,0.036580443382263184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,768,0.030555556217829388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,1536,0.03541422221395705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,1024,0.034068445364634194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,65536,0.3077804512447781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,16384,0.08041155338287354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,16384,0.04864177770084805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,12288,0.06708088848325941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,65536,0.1842728853225708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,12288,0.03924177752600776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,10240,0.05634755558437771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,10240,0.04698933495415581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,8192,0.026871111657884385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,8192,0.05105422271622551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,256,0.02907022171550327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,128,0.027279110418425664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,7168,0.04436799883842468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,7168,0.036652445793151855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,6144,0.021430222524536982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,6144,0.039311110973358154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,512,0.03087466624048021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,5120,0.02567822237809499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,5120,0.03414933217896356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,16384,0.10458133618036906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,4096,0.02163733376397027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,4096,0.030270222160551283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,12288,0.08739555544323391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,3584,0.01385600037044949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,3584,0.028586665789286297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,65536,0.33518134223090273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,3072,0.014565333724021912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,3072,0.02700888944996728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,10240,0.07764710982640584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,2560,0.012466666599114736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,2560,0.024845333562956914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,8192,0.06653866502973768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,2048,0.010513777534166971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,2048,0.02299555473857456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,7168,0.06424444251590304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,1536,0.008461332983440822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,1536,0.020505777663654752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,6144,0.05721777677536011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,1024,0.006694222076071634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,1024,0.018584888842370775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,5120,0.050642665889528066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,768,0.005688000056478713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,768,0.017850667238235474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,3584,0.04275200102064344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,512,0.0046942221621672315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,512,0.016914667354689706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,4096,0.04435822367668152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,256,0.003952888978852166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,256,0.01575377749072181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,3072,0.04155022237035964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,2560,0.03983911209636264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,128,0.003664888855483797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,128,0.015231110983424716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,64,0.003408000080121888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,64,0.015223999818166098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,32,0.003604444364706675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,32,0.015254222684436373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,2048,0.03637955586115519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,65536,0.30156445503234863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,65536,0.1600666708416409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,16384,0.04622933268547058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,768,0.03072355522049798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,1536,0.03500888744990031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,16384,0.07826133569081624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,12288,0.06464799907472399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,12288,0.0481608874268002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,1024,0.03307111064592997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,10240,0.055642664432525635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,10240,0.029322667254341975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,8192,0.030568887790044148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,8192,0.04747200012207031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,128,0.027424888478385076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,256,0.028817776176664565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,7168,0.04181866513358223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,7168,0.021749332547187805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,6144,0.02141777839925554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,6144,0.03876533442073398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,512,0.03030666708946228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,5120,0.016488000750541687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,5120,0.03401511245303684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,65536,0.40830132696363663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,4096,0.013375110924243927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,4096,0.029903110530641343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,16384,0.11180355813768174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,3584,0.012153777811262341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,3584,0.028505778974956934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,12288,0.09133955505159165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,3072,0.01352266636159685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,3072,0.026557332939571802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,10240,0.08047555552588569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,2560,0.011379555695586734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,2560,0.024907555845048692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,8192,0.0697386662165324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,2048,0.009408888717492422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,2048,0.023303111394246418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,7168,0.06500889195336236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,1536,0.0081377774477005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,1536,0.020067556036843192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,6144,0.059783114327324756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,1024,0.006206222292449739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,1024,0.0184897780418396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,5120,0.052800890472200185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,768,0.0052844443255000645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,768,0.017699556218253244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,4096,0.046612444851133555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,512,0.004471999903519948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,512,0.016807110773192514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,3584,0.04381155636575487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,256,0.003926222109132343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,256,0.015842666228612263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,3072,0.04319999946488274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,128,0.003593777616818746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,128,0.015207999282413058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,2560,0.040589332580566406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,64,0.003371555565132035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,64,0.015188443991872998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,32,0.0035173334181308746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,32,0.015036443869272867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,2048,0.037307557132509016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,65536,0.2932373417748345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,1536,0.03420977791150411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,65536,0.1415191094080607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,1024,0.03247288862864176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,768,0.03019555409749349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,16384,0.042782220575544566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,12288,0.06112800041834513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,16384,0.07696355713738336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,12288,0.02961777647336324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,10240,0.05393422312206692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,10240,0.028367999527189467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,8192,0.023288889063729182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,8192,0.04640266630384657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,512,0.03074044320318434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,7168,0.02085777786042955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,7168,0.04182666540145874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,256,0.02903466754489475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,6144,0.018588445252842374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,128,0.027678221464157104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,6144,0.03827555643187629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,5120,0.013824889229403602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,5120,0.033759110503726535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,16384,0.08770133389366998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,65536,0.2821902169121636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,4096,0.01351377781894472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,12288,0.07317422495947944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,4096,0.030047112041049536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,3584,0.0107022225856781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,3584,0.028254220883051556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,3072,0.009725333087974125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,3072,0.026671111583709717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,10240,0.06532088915506999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,2560,0.00999022192425198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,2560,0.02420888841152191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,8192,0.05743288993835449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,2048,0.008548444343937768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,2048,0.021906667285495337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,7168,0.054042667150497437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,1536,0.00665244460105896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,1536,0.019837333096398246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,6144,0.049944000111685864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,1024,0.005146666533417172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,1024,0.018087110585636564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,5120,0.045666668150160045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,768,0.004888888862397936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,768,0.017332444588343304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,4096,0.040695998403761126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,512,0.004097777936193678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,512,0.01665422154797448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,3584,0.039990223116344877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,256,0.00370133337047365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,256,0.015543111496501498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,3072,0.03908266623814901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,128,0.0034168888297345904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,128,0.014974221587181091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,2560,0.03749155667093065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,64,0.0032213332338465583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,64,0.015007111761305066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,2048,0.03491200009981791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,32,0.0032666667054096856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,32,0.014927110738224454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,65536,0.29133423169453937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,65536,0.11602311664157444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,1536,0.032270222902297974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,16384,0.03972533345222473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,16384,0.07588355408774482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,1024,0.031581332286198936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,12288,0.06195022000206841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,12288,0.03165600034925673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,10240,0.028704888290829126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,10240,0.05339466532071432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,768,0.03036266565322876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,8192,0.01910044418440925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,8192,0.04624888963169522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,512,0.03019022279315525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,7168,0.017257778180970084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,7168,0.041351109743118286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,256,0.02857066525353326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,128,0.027088888817363318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,6144,0.015263110399246216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,6144,0.03791911072201199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,5120,0.01735377808411916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,5120,0.033533334732055664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,65536,0.2760302225748698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,4096,0.014628445108731588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,4096,0.03014844324853685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,12288,0.07214844226837158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,3584,0.012831110921170978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,3584,0.028102222416136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,16384,0.08576177888446385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,3072,0.011798222031858234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,3072,0.026320000489552815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,10240,0.06316088967853122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,2560,0.010168888502650792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,2560,0.023710222707854375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,8192,0.056088891294267446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,2048,0.008810666700204214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,2048,0.02273511058754391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,7168,0.05286666750907898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,1536,0.007139555282062954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,1536,0.019825778073734708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,5120,0.04486933350563049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,6144,0.048700445228152804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,1024,0.004927111168702443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,1024,0.018036444981892902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,768,0.004354666504595014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,768,0.017256889078352187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,4096,0.039860443936453924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,512,0.003991111285156674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,512,0.0165964447789722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,3584,0.03999733262591892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,256,0.003649777836269803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,256,0.01566844516330295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,3072,0.038683556848102145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,128,0.0033208889265855155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,128,0.015000000596046448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,2560,0.03502755694919162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,64,0.003112888791494899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,2048,0.03396888905101352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,64,0.01496444476975335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,32,0.0033208889265855155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,32,0.014729777971903482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,65536,0.0941164427333408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,1536,0.031798223654429116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,65536,0.29062755902608234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,16384,0.07592266798019409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,16384,0.029899554120169744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,1024,0.031164444155163232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,12288,0.025117332736651104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,768,0.029355555772781372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,12288,0.06134577592213949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,10240,0.022209778428077698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,10240,0.05303288830651177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,8192,0.01832177738348643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,8192,0.04498044318623013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,512,0.029858665333853826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,7168,0.016520000166363187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,256,0.028833776712417603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,7168,0.04171289006868998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,6144,0.014750222365061441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,6144,0.037929776642057635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,128,0.0270506673389011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,5120,0.013068444199032254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,5120,0.03336000111367967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,16384,0.08308800061543782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,4096,0.011078221930397881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,4096,0.02966044346491496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,65536,0.2735511196984185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,3584,0.010125333236323463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,3584,0.028005331754684448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,12288,0.06961511241065131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,3072,0.00923288861910502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,10240,0.062034666538238525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,3072,0.025640888346566096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,2560,0.008224888808197444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,8192,0.054217778974109225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,2560,0.023360000716315374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,2048,0.007264888948864407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,2048,0.02145066691769494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,7168,0.05177333288722568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,1536,0.006090666684839461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,1536,0.019426667028003268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,6144,0.04884355598025852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,1024,0.004859555512666702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,1024,0.017994667092959087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,5120,0.044771555397245616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,768,0.004547555413511065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,768,0.01731911136044396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,4096,0.039694223139021136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,3584,0.03953244288762411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,512,0.003899555653333664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,512,0.016514667206340365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,256,0.0035013332962989807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,256,0.015509333875444202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,3072,0.037717332442601524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,2560,0.036269333627488874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,128,0.0033031110134389666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,128,0.014967110421922473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,64,0.003144888828198115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,2048,0.03235466612709893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,32,0.003144888828198115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,64,0.014856888188256158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,32,0.014887111054526435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,65536,0.28964000278049046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,65536,0.06559822294447157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,16384,0.025763556361198425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,16384,0.07604710923300849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,1536,0.03219733304447598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,12288,0.019536000159051683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,12288,0.06113333172268338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,1024,0.031067556805080835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,10240,0.01627288924323188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,768,0.02937155630853441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,10240,0.052547554175059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,8192,0.014122666584120857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,8192,0.044807109567854136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,256,0.02849422229660882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,512,0.029819554752773706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,7168,0.01311555587583118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,7168,0.040595554643207125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,6144,0.012176888684431711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,6144,0.0376506679587894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,128,0.02680355476008521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,5120,0.010834666589895884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,5120,0.03282577792803446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,65536,0.3508293363783095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,4096,0.009377777576446533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,4096,0.02877955635388692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,16384,0.09992444515228271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,3584,0.008747555315494537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,3584,0.026685333914226953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,12288,0.08239466614193387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,3072,0.00795999997191959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,3072,0.02453600035773383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,8192,0.06177066432105171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,2560,0.007338666253619724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,10240,0.07189510928259955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,2560,0.02295644415749444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,2048,0.006455111006895701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,2048,0.02142933342191908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,7168,0.059472892019483775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,1536,0.005395555661784277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,1536,0.01962844365172916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,6144,0.05459910962316725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,1024,0.004527111020353106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,1024,0.017893332574102614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,5120,0.048650665415657886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,768,0.00416711096962293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,768,0.017086222767829895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,4096,0.04324088825119866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,512,0.003735110991530948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,512,0.016544888416926067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,3584,0.040758222341537476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,256,0.0033982222278912864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,256,0.015391111373901367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,3072,0.038725333081351385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,128,0.0032213332338465583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,128,0.015020444989204407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,2560,0.036369777388042875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,64,0.0029502221279674103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,64,0.014872888724009195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,32,0.003018666679660479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,32,0.01458044515715705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,65536,0.051479998562071055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,2048,0.033744888173209295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,65536,0.2908079889085558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,16384,0.018375999397701688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,16384,0.07604266537560357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,1536,0.032940444019105695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,12288,0.017501332693629794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,12288,0.061179558436075844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,1024,0.031495110856162176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,10240,0.015616888801256815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,10240,0.05145688851674398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,768,0.029672000143263075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,8192,0.013777777552604675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,8192,0.044050667021009654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,512,0.030193779203626845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,7168,0.01369599997997284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,7168,0.04132533404562209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,256,0.0288426677385966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,6144,0.012359110845459832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,6144,0.0376853346824646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,128,0.02741333345572154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,5120,0.012503999802801343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,5120,0.03250311149491204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,65536,0.3180595503913032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,4096,0.009475555684831407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,16384,0.09679644637637669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,4096,0.028724445237053767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,3584,0.008795555267069075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,3584,0.02681422233581543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,12288,0.0797591143184238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,3072,0.007958222594526079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,3072,0.024992000725534227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,10240,0.0699839989344279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,2560,0.0069155556460221606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,2560,0.022925333844290838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,8192,0.060565332571665444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,2048,0.006016888966162999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,2048,0.021154666940371197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,7168,0.05829333596759372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,1536,0.00526311124364535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,1536,0.01974933346112569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,6144,0.05348444316122267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,1024,0.004602666530344221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,1024,0.01789777808719211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,5120,0.04740355412165324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,768,0.004232888834344016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,768,0.017140444782045152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,4096,0.042797333664364286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,512,0.003968888686762916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,512,0.016608888904253643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,3584,0.039509332842297025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,256,0.0034142221427626083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,256,0.01555999947918786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,3072,0.039956443839603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,128,0.0031617778456873367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,128,0.014848000473446317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,2560,0.035792890522215105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,64,0.0029582221888833572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,64,0.0148053334818946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,32,0.0029253332565228143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,32,0.014630221658282809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,2048,0.033088889386918806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,65536,0.04055288765165541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,65536,0.28774576716952854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,16384,0.014255111416180929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,16384,0.0754879977968004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,1536,0.032504889700147845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,12288,0.013561777770519257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,12288,0.059531556235419385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,1024,0.031141334109836157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,10240,0.01255644443962309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,10240,0.051516443490982056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,768,0.029995554023318823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,8192,0.011167111496130625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,8192,0.04437511165936788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,256,0.028713779317008123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,512,0.029996444781621296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,7168,0.010715555813577441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,7168,0.04046577877468533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,6144,0.009961778091059791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,6144,0.03682933251063029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,128,0.027264000640975103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,5120,0.00964355551534229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,5120,0.032207998964521624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,65536,0.2578168974982368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,4096,0.008276444342401292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,4096,0.027785778045654297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,16384,0.07840889030032687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,3584,0.007844444778230455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,3584,0.026133333643277485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,12288,0.06532000170813666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,3072,0.007081777685218387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,3072,0.024711999628278945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,10240,0.05808977948294746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,2560,0.006434666613737742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,2560,0.0228426671690411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,8192,0.05141689048873054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,2048,0.005708444449636672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,2048,0.020794666475719877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,7168,0.04983466532495287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,1536,0.005010666532648934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,1536,0.01945688823858897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,6144,0.046675556235843234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,1024,0.004287111262480418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,1024,0.01779822177357144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,5120,0.04143288731575012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,768,0.003941333128346337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,768,0.01698755555682712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,4096,0.036291556225882635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,512,0.003585777762863371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,512,0.016177778442700703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,3584,0.035651554663976036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,3072,0.035024000538720026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,256,0.0032879999942249725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,256,0.015507555670208402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,128,0.00309244439833694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,128,0.0148044443792767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,2560,0.03334666623009576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,64,0.002926222152180142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,64,0.014673777752452426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,32,0.0029582221888833572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,32,0.014559111661381192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,65536,0.03544355432192484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,65536,0.29131465488009983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,2048,0.03171022070778741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,16384,0.015000000596046448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,16384,0.07469155391057332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,1536,0.030638221237394545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,12288,0.011912888950771756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,12288,0.05996711386574639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,1024,0.03046222196684943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,10240,0.011190222369299995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,10240,0.05103466577000088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,768,0.029616001579496596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,8192,0.009926222264766693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,8192,0.04370933439996508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,512,0.029496888319651287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,7168,0.009039999710188972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,7168,0.03927288783921136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,128,0.026928888426886663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,6144,0.008186666501892937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,256,0.02815999918513828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,6144,0.03542133172353109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,5120,0.007682667010360294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,5120,0.03130844566557143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,65536,0.25708622402615017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,4096,0.007194666398896112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,4096,0.027627555860413447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,16384,0.07840889030032687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,3584,0.007526222202512953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,3584,0.025920000341203477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,12288,0.06419822242524889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,3072,0.007148444652557373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,3072,0.02427022159099579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,10240,0.05751644240485298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,2560,0.0063999998900625445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,2560,0.022475555539131165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,8192,0.05145600106981066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,2048,0.005667555663320754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,2048,0.020853334003024634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,7168,0.04929600159327189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,1536,0.00499288903342353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,1536,0.01904000010755327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,5120,0.040568000740475126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,1024,0.004304000073009067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,6144,0.04378933376736111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,1024,0.01751733322938283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,768,0.003965333518054751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,768,0.016792888442675274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,4096,0.035769777165518865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,512,0.0035617777870761026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,512,0.01590755581855774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,3584,0.03591466612286038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,256,0.0032177778581778207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,256,0.01538133290078905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,3072,0.03467199868626065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,128,0.0030248889492617715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,2560,0.03314844436115689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,128,0.014924445086055331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,64,0.0028817777832349143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,64,0.014866666661368476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,32,0.0029111111329661477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,32,0.014864888456132678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,65536,0.02609866691960229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,65536,0.28926311598883736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,2048,0.03115288747681512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,16384,0.010052444206343757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,16384,0.07477333148320515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,1024,0.03012711140844557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,1536,0.030540443129009668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,12288,0.00848711116446389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,768,0.029155555698606703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,10240,0.00795199970404307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,12288,0.05951733059353299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,10240,0.05032177766164144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,8192,0.007479999628331926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,8192,0.042211555772357516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,512,0.029492444462246362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,7168,0.007324444750944774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,7168,0.038592888249291316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,256,0.02794933319091797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,6144,0.0069120000633928515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,6144,0.035169776942994856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,128,0.026775999201668635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,5120,0.007393777370452881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,5120,0.03095199995570713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,65536,0.25661688380771214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,4096,0.006919999917348226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,4096,0.027440889014138117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,16384,0.07787377966774835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,3584,0.007360888852013483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,3584,0.025807999902301367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,10240,0.05695733096864489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,12288,0.06426933076646593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,3072,0.0069671107663048645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,3072,0.02395733363098568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,2560,0.006188444379303191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,2560,0.022220444348123338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,8192,0.04894222153557671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,2048,0.005519111123349931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,2048,0.020657777786254883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,7168,0.0468444459968143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,1536,0.004866666677925321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,1536,0.01901244454913669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,6144,0.04357599880960253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,1024,0.004208000169859992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,1024,0.01721955504682329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,5120,0.03982933362325033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,768,0.0038240001433425476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,768,0.016537777251667447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,4096,0.03616888986693488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,512,0.003555555724435382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,512,0.015881778465376962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,3584,0.0357528891828325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,256,0.0031679999083280563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,256,0.01515111161602868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,3072,0.03463822272088792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,128,0.002983999955985281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,128,0.014550222290886773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,2560,0.033152888218561806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,64,0.002961777771512667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,2048,0.031176888280444678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,64,0.014461333552996317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,32,0.0028435554769304064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,32,0.014202665951516895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,65536,0.014720888601409065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,16384,0.008007999923494127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,16384,0.0743866695298089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,65536,0.29048087861802846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,12288,0.007368000017272101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,10240,0.007277333074145847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,12288,0.058096885681152344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,8192,0.007174222005738153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,10240,0.04948000113169352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,7168,0.0069955554273393415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,8192,0.041512889994515315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,6144,0.00684533350997501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,7168,0.0380364457766215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,5120,0.007230222225189209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,6144,0.03471644388304816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,4096,0.006778666542636023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,5120,0.03085600005255805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,3584,0.007253333098358578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,4096,0.027274666561020747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,3072,0.0069253332912921906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,1536,0.030368000268936157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,3584,0.025711999999152288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,2560,0.006199999815887875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,3072,0.023930667175187006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,2560,0.022282666630215112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,2048,0.005515555540720622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,1536,0.004836444639497333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,2048,0.020474667350451153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,1024,0.004137777619891697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,1536,0.01891288823551602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,768,0.0038088887102074097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,1024,0.017115554875797696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,512,0.003481777591837777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,768,0.01648977730009291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,256,0.003160888950030009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,512,0.015670221712854173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,128,0.002941333378354708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,256,0.015138667490747241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,64,0.002814222127199173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,128,0.014440889159838358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,32,0.0027644443843099806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,64,0.01444355481200748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,32,0.014203555054134794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,65536,0.013633777697881063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,16384,0.007083555890454187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,16384,0.07428889142142402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,65536,0.290704886118571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,12288,0.006984000404675801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,10240,0.00721244431204266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,12288,0.05908177958594429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,8192,0.007061333292060428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,10240,0.04941244588957893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,7168,0.006973333656787872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,8192,0.04133244355519613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,6144,0.00674222202764617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,7168,0.03801777627733018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,5120,0.007245333658324347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,6144,0.03472266594568888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,4096,0.006688888702127669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,5120,0.030881779061423406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,3584,0.007160888777838812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,4096,0.027277333868874445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,3072,0.006832888970772426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,3584,0.02569599946339925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,2560,0.006128888991143968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,3072,0.023901333411534626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,2048,0.005453333258628845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,2560,0.022286223040686712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,1536,0.00479911102188958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,2048,0.020464888877338834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,1024,0.004159111115667555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,1536,0.018910222583346896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,768,0.0037662221325768363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,1024,0.017142222987280954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,512,0.0034622223012977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,768,0.01645511057641771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,1024,0.030179556873109605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,256,0.0031600000543726813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,512,0.015678221980730694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,128,0.0029564443975687027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,256,0.015135111080275642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,64,0.002814222127199173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,128,0.014482667048772177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,32,0.00282577777074443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,64,0.014455111490355598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,32,0.01423377792040507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,768,0.02872177627351549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,512,0.0295688874191708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,256,0.02797777785195245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,128,0.026739555928442214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,16384,0.5934649043613011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,12288,0.4518204265170627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,10240,0.3944408893585205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,12288,0.5045617951287164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,16384,0.609958224826389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,10240,0.4926737679375543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,12288,0.5888524585300022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,8192,0.32371022966172963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,10240,0.42479289902581113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,7168,0.28411822848849827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,8192,0.35214487711588544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,7168,0.3146257665422228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,6144,0.24723733796013725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,8192,0.39455199241638184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,5120,0.20927466286553276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,6144,0.25682221518622506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,16384,0.7750062412685819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,6144,0.30336443583170575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,5120,0.22871200243631998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,7168,0.3478417661454942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,4096,0.16767822371588814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,3584,0.15085244178771973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,4096,0.18575467003716362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,5120,0.25485867924160427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,3584,0.16691199938456217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,3072,0.1277244488398234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,4096,0.2065075503455268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,3072,0.14880443943871394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,3584,0.18143110805087617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,2560,0.11008266607920329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,2048,0.09455022546980117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,2560,0.13329511218600804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,2560,0.12926933500501844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,3072,0.15799110465579563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,2048,0.10989599757724339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,2048,0.10923377672831218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,1024,0.04937155710326301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,1536,0.07060177458657159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,1024,0.07099999984105428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,1536,0.08873866664038764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,768,0.0428017775217692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,1024,0.06110755602518717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,768,0.06196977694829305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,768,0.0483857790629069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,256,0.03014755580160353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,256,0.022150221798155043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,512,0.03622311022546556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,512,0.053561776876449585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,128,0.013384000294738345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,1536,0.08588533269034491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,512,0.03731911049948798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,64,0.010455999937322406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,128,0.027514666318893433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,256,0.045758220884535045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,64,0.029191109869215224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,32,0.01054666688044866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,128,0.04050577680269877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,32,0.03313066562016805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,16384,0.17386666933695474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,12288,0.13557422161102295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,16384,0.20402399698893228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,12288,0.14682311481899685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,16384,0.19118044111463758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,10240,0.11912800206078424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,12288,0.16230132844713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,10240,0.14278311199612087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,10240,0.12404266993204753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,8192,0.09859644704394871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,8192,0.09938489066229926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,8192,0.11628977457682292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,7168,0.08845866388744777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,7168,0.10650844706429376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,6144,0.07692178090413411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,7168,0.08750310871336196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,6144,0.08934311072031657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,65536,0.6575857798258463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,5120,0.06556889083650377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,6144,0.09348444143931071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,4096,0.056537780496809215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,5120,0.08335466517342462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,5120,0.0655075576570299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,4096,0.053676442967520825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,4096,0.07064266337288751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,3584,0.051389333274629384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,3584,0.04772888951831394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,3072,0.04541155695915222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,3584,0.06660177972581652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,3072,0.06222222248713175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,3072,0.0422071119149526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,2560,0.03640888796912299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,2560,0.04053599966896905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,2560,0.05710311068428887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,65536,0.7040008968777127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,1536,0.031160887744691636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,1024,0.024382222029897902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,1536,0.04772177669737074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,2048,0.030299554268519085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,2048,0.036271111832724676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,768,0.013372444444232516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,1024,0.04172711239920723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,512,0.010068444742096795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,768,0.022361778550677832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,2048,0.05234666665395101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,1536,0.024519999821980793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,512,0.020963556236690942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,768,0.03979911075698005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,256,0.007303111255168915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,512,0.03714399867587619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,256,0.018374222848150466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,128,0.005162666655249066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,64,0.0046568889584806235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,256,0.0339680016040802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,128,0.016718222035302054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,32,0.005055111315515307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,128,0.029613332615958318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,64,0.01720177796151903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,1024,0.017755554782019723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,32,0.017864889568752713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,16384,0.146814227104187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,16384,0.15523110495673284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,12288,0.130185776286655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,12288,0.11423467265235053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,12288,0.12530222203996447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,16384,0.16565867265065512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,10240,0.09490666786829631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,10240,0.0994231104850769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,65536,0.7985226843092176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,10240,0.10773244169023301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,8192,0.08565244409773086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,8192,0.07636621925565931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,7168,0.07666310999128553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,8192,0.09178755680720012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,7168,0.08447377549277411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,7168,0.07896266380945842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,6144,0.06633333365122478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,6144,0.0746506651242574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,5120,0.058008001910315625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,5120,0.06694400310516357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,6144,0.05995111332999336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,5120,0.05113955669932895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,4096,0.04850755466355217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,65536,0.5742115444607204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,4096,0.058535999721950956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,3584,0.045553776952955455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,4096,0.041323555840386286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,3584,0.03671288821432326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,3072,0.040403554836908974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,3584,0.05814222494761149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,2560,0.0370124446021186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,3072,0.03258400162061056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,3072,0.05391022231843737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,2560,0.04901511139339871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,2048,0.03206933206982083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,2048,0.0461404456032647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,1536,0.02790488799413045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,1024,0.01276088919904497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,2048,0.02342044479317135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,1536,0.04329866502020094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,1024,0.022919111781650122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,2560,0.03137600090768602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,768,0.010638222098350525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,1536,0.01869155632125007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,768,0.021920889616012573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,1024,0.038429333104027644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,512,0.00849955528974533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,65536,0.5461040072970921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,256,0.0058702222175068325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,512,0.020390222469965618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,768,0.037913779417673744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,128,0.00435644429590967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,512,0.036271999279658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,256,0.017804443836212158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,128,0.016504888733228047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,64,0.003935111065705617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,256,0.031189332405726116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,32,0.004317333300908406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,128,0.028964443339241877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,64,0.01680888897842831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,32,0.017135999268955655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,16384,0.14362933900621203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,12288,0.099053336514367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,16384,0.148199995358785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,12288,0.11232533719804551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,65536,0.5981955528259277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,16384,0.14903910954793295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,12288,0.11756977770063613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,10240,0.09785155455271403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,8192,0.08041777875688341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,10240,0.08252088891135322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,10240,0.10312889019648235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,8192,0.08742133114073013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,7168,0.07223377625147502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,8192,0.06636089086532593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,7168,0.08340622319115533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,7168,0.059736000166998975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,6144,0.0524453322092692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,6144,0.06270933151245117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,5120,0.05408711234728495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,6144,0.07128800286187066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,5120,0.04452355702718099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,5120,0.06424888637330797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,4096,0.04680266645219591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,4096,0.05654666821161906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,4096,0.03534133235613505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,3584,0.0433937774764167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,3584,0.054470221201578774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,3584,0.03202133377393087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,3072,0.03971733318434821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,65536,0.5487964418199327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,3072,0.05170311199294197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,2560,0.0351457761393653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,2048,0.03049955434269375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,2560,0.04696266518698799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,2048,0.04359022114011976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,1536,0.016028443972269695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,1536,0.027306665976842243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,3072,0.02864533331659105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,1024,0.01163822246922387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,2560,0.024531554844644334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,1536,0.04212622178925408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,1024,0.02272266646226247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,2048,0.02050044470363193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,768,0.009776888622177972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,512,0.00796977761718962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,1024,0.03780000077353583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,768,0.021566222111384075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,768,0.03760533200369941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,256,0.005817777580685086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,512,0.019155555301242404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,512,0.034833778937657676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,128,0.004587555511130227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,256,0.017536888519922893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,256,0.030253334177864924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,128,0.01646577815214793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,64,0.004032888760169347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,128,0.028618666860792372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,32,0.0042088888585567474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,64,0.016672000288963318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,32,0.016932444439993966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,65536,0.556389331817627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,16384,0.10720622539520264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,16384,0.12067643801371257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,12288,0.08490044540829128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,16384,0.09790044360690647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,12288,0.07550488577948676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,65536,0.5147537655300564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,12288,0.09975110822253758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,10240,0.07464000251558092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,10240,0.06407733096016778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,8192,0.062000003125932485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,10240,0.08932711018456353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,8192,0.051755554146236844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,65536,0.38974311616685653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,7168,0.05618311299218071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,7168,0.0729279981719123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,8192,0.07750578059090509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,7168,0.046033776468700834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,6144,0.046663112110561795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,6144,0.04909955461819967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,5120,0.043228444125917226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,6144,0.06561511092715792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,5120,0.05925066603554619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,5120,0.03493600090344747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,4096,0.03784977727466159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,3584,0.03558844327926636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,3584,0.05016533202595181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,3072,0.03231733375125461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,4096,0.05155466662512886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,4096,0.028724445237053767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,2560,0.02946399980121189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,3072,0.04667377803060743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,2560,0.04346488912900289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,3584,0.025721778472264607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,3072,0.022851554883850947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,2048,0.02604622145493825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,1536,0.0125591109196345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,2048,0.04104177819357978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,2560,0.019638222124841478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,1024,0.009594666461149851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,1536,0.023583110835817125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,1536,0.039846221605936684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,1024,0.020649777518378366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,768,0.008178667061858708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,2048,0.016159110599093966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,1024,0.036171555519104004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,768,0.019121777680185106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,512,0.0067528887755341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,768,0.036406222316953875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,512,0.032912890116373696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,512,0.01790488925245073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,256,0.004353777815898259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,256,0.016361777981122334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,128,0.015497777197096082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,128,0.0038897775941424896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,64,0.0035573335157500375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,256,0.029792000850041706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,32,0.00398577791121271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,128,0.028510222832361858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,64,0.0155102229780621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,32,0.015636444091796875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,65536,0.4144906732771132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,16384,0.10334399673673843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,16384,0.13426132996877035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,12288,0.06775555345747206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,12288,0.0810862249798245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,12288,0.10730222198698257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,16384,0.08586666981379192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,10240,0.07069155242707995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,10240,0.09554933177100287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,8192,0.059417777591281466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,10240,0.05691288577185737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,8192,0.0455022222465939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,8192,0.08061155345704821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,65536,0.41495288742913145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,65536,0.3604631159040663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,7168,0.05455466773774889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,7168,0.04127466678619385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,7168,0.07548711034986708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,6144,0.04872533347871569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,6144,0.06655999686982897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,5120,0.04218399855825636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,4096,0.036581334140565656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,6144,0.036672890186309814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,4096,0.05316266748640272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,3584,0.03391289048724704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,5120,0.06033511294258965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,5120,0.031005332867304485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,3584,0.05052355594105191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,3072,0.0318195554945204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,3072,0.04710666669739617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,4096,0.025432889660199482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,3584,0.02308977809217241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,2560,0.02836444311671787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,3072,0.02066044509410858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,2048,0.014293332894643148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,2560,0.0429351098007626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,2048,0.025232000483406916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,1536,0.01148622234662374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,2048,0.04184533490075005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,1536,0.02300888962215847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,1024,0.00886933339966668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,2560,0.01777777738041348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,1024,0.02031822171476152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,1536,0.03881599836879306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,1024,0.036791112687852644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,768,0.018168000711335074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,768,0.007636444436179266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,512,0.006454222318198945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,768,0.03580800029966567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,512,0.017681777477264404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,256,0.004316444612211651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,512,0.03292355603641934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,256,0.016326222154829238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,128,0.0038764443662431505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,256,0.029826667573716905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,128,0.015270221564504834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,64,0.0038684445122877755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,65536,0.47559912999471027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,32,0.0039048890272776284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,128,0.028042667441897925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,64,0.01551555593808492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,32,0.01540177729394701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,16384,0.09880710972679986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,16384,0.10575644175211589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,65536,0.35650311575995547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,16384,0.07695733176337348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,12288,0.07718577649858263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,12288,0.08633510933982001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,10240,0.06628533204396565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,12288,0.06020000245836046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,10240,0.050507555405298867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,10240,0.077528887324863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,8192,0.05644622113969591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,8192,0.04084088736110263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,8192,0.06819111108779907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,7168,0.03695377707481384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,7168,0.052292442984051175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,7168,0.06457422176996867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,5120,0.040361778603659734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,6144,0.032791998651292585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,6144,0.04572622312439812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,5120,0.05285511083073086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,4096,0.03520355621973673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,6144,0.060030221939086914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,3584,0.03293777836693658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,4096,0.047035555044809975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,5120,0.0281315545241038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,4096,0.02276977731121911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,65536,0.3450853294796414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,65536,0.34656355116102433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,3584,0.04584088921546936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,3072,0.03014933400683933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,2560,0.015613334046469795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,2560,0.02780888809098138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,3072,0.04371644390953911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,2048,0.012983111043771109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,3584,0.02087200019094679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,2048,0.024688889582951862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,1536,0.010624888870451184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,2560,0.04112622141838074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,3072,0.018570666511853535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,2048,0.039565334717432656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,1536,0.022602667411168415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,1024,0.008413333031866286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,1536,0.038366221719317965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,1024,0.02012622190846337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,768,0.007311999797821045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,768,0.0181688881582684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,1024,0.03609866566128201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,512,0.005778666585683823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,768,0.03602222270435757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,256,0.004353777815898259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,512,0.01737422247727712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,512,0.031826668315463595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,256,0.016209777858522203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,128,0.003941333128346337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,256,0.029291556941138372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,128,0.015411555767059326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,64,0.0035706667436493766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,32,0.0037333332002162933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,64,0.015190222197108798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,128,0.027817779117160376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,32,0.015352000792821249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,16384,0.08987644645902847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,16384,0.12438133027818467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,12288,0.05297955539491442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,12288,0.06977155473497179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,16384,0.07482044564353095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,12288,0.09956622123718262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,10240,0.04467022087838915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,10240,0.060991108417510986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,8192,0.052143110169304736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,10240,0.0889164474275377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,8192,0.03614577651023865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,8192,0.0754515528678894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,7168,0.04857777886920505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,7168,0.07173066669040255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,65536,0.3088346587287055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,7168,0.03276800115903219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,6144,0.04249777727656894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,6144,0.029112001260121662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,5120,0.037300444311565824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,6144,0.06247466802597046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,5120,0.05708177884419759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,4096,0.03316888875431485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,65536,0.3207804361979167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,4096,0.050415999359554715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,4096,0.020626667473051283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,3584,0.029860443539089624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,5120,0.02495288848876953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,3072,0.016201777590645682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,3584,0.04785155587726169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,3072,0.02725155486000909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,2560,0.013883555928866068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,3584,0.018769777483410306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,2560,0.025613332788149517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,3072,0.04513955447408888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,2048,0.01180355581972334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,1536,0.009878222313192155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,2560,0.04182044333881802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,2048,0.02362577782736884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,1536,0.02166844407717387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,2048,0.03964266512129042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,1024,0.008110221889283922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,1024,0.019069333871205647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,768,0.0070186663005087115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,1536,0.03804355528619554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,1024,0.03515377640724182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,768,0.017966222431924608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,65536,0.43852443165249294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,512,0.005559111220969095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,256,0.004303110970391167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,512,0.017236444685194228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,768,0.03426400158140395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,512,0.031238221459918555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,256,0.02922933300336202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,128,0.0038364442686239877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,256,0.016026667422718473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,128,0.015384000208642749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,64,0.0034942221310403612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,128,0.027824888626734417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,32,0.0035822221802340615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,64,0.01516355574131012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,32,0.015075556106037565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,16384,0.08367466926574707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,65536,0.24940710597568086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,16384,0.10105866856045193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,12288,0.06485066811243693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,16384,0.05905066596137153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,12288,0.0885573360655043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,12288,0.04821066723929512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,10240,0.056609776284959584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,10240,0.04189866781234741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,8192,0.049435555934906006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,10240,0.07522400220235188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,8192,0.03551822238498264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,8192,0.0654311113887363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,7168,0.045520000987582736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,65536,0.30711200502183705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,6144,0.038367112477620445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,7168,0.06219910913043552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,6144,0.05488711264398363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,5120,0.03474844495455424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,5120,0.04971999923388163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,7168,0.03380533390574985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,4096,0.02991466720898946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,6144,0.029354665014478896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,3584,0.014184888866212634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,5120,0.020190222395790946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,4096,0.04455022348297966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,3584,0.02858933475282457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,3072,0.013311111264758639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,3584,0.04269688990381029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,4096,0.015839111473825242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,3072,0.026495999760097925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,2560,0.01090311093462838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,3072,0.04122311207983229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,2048,0.009364444348547194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,2560,0.0248355550898446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,2560,0.038238220744662814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,2048,0.02316355539692773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,65536,0.3140960004594591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,1536,0.008275555239783393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,2048,0.0372995568646325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,1536,0.021116443806224402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,1024,0.006575999988449945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,1536,0.03638933433426751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,1024,0.01860000027550591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,768,0.005763555566469829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,768,0.031785776217778526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,768,0.01774488886197408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,1024,0.03239733311865065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,512,0.004519111166397731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,512,0.017059556312031217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,256,0.003943110919660992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,512,0.03022577696376377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,256,0.016040888097551134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,128,0.003573333223660787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,256,0.028743998871909246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,128,0.01514488955338796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,64,0.0033324443631701996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,32,0.0034862222770849862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,128,0.027259555127885606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,64,0.015060444672902426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,32,0.015023999743991427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,16384,0.07990933126873441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,16384,0.09844444195429485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,12288,0.06230755647023519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,16384,0.05214577913284302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,12288,0.04317511121431986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,10240,0.054618666569391884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,12288,0.08184533649020724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,65536,0.2044079966015286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,8192,0.04699288805325826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,10240,0.037688887781567044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,8192,0.0643448895878262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,10240,0.07353688610924615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,7168,0.04268266757329305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,7168,0.06056444512473213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,8192,0.025616889198621113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,6144,0.037419554260041975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,7168,0.023457777169015672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,65536,0.2751137680477566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,6144,0.053377777338027954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,6144,0.021015110943052504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,5120,0.033896889951494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,4096,0.014380445082982382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,4096,0.02983022067281935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,5120,0.017833777599864535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,5120,0.04882488979233635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,3584,0.012864000267452665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,4096,0.043431109852261014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,3584,0.02791555391417609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,3072,0.011481777661376528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,3584,0.0420702232254876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,3072,0.026126222478018865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,2560,0.01016533292002148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,2560,0.02453244394726223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,2048,0.008809777597586313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,3072,0.04059111078580221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,2560,0.03790488839149475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,2048,0.02292622294690874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,1536,0.007656888829337225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,65536,0.30644888348049587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,2048,0.037167112032572426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,1536,0.020996444755130347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,1024,0.006088888893524806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,1024,0.01848355597919888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,1536,0.03605777687496609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,768,0.004935111022657818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,1024,0.03383822242418925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,768,0.017654221918847825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,512,0.004137777619891697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,768,0.03147644466824002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,512,0.01683555543422699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,256,0.003760888758632872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,256,0.02875466810332404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,512,0.0302817788388994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,256,0.015974221958054435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,128,0.0034097778714365433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,128,0.01501066651609209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,64,0.0036880001425743103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,64,0.014926221635606555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,32,0.003504000190231535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,128,0.027269333600997925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,32,0.0151573336786694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,65536,0.17793688509199354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,16384,0.07760800255669488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,12288,0.06017866399553087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,16384,0.11042844586902195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,10240,0.052813334597481616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,16384,0.04652888907326592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,12288,0.03863644599914551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,8192,0.04494222336345249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,10240,0.08153155777189466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,12288,0.0912062194612291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,8192,0.0699146654870775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,10240,0.03352177805370755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,8192,0.02881866693496704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,7168,0.040976888603634305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,7168,0.06583378050062391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,6144,0.03637955586115519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,7168,0.023718222975730896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,6144,0.05765866571002536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,5120,0.019315555691719055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,5120,0.03268266717592875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,6144,0.02112355497148302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,4096,0.014177777700954013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,4096,0.029045333464940388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,5120,0.05276977684762743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,3584,0.011938666303952536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,4096,0.046894222497940063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,3584,0.027870221270455256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,3072,0.010861333045694562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,3584,0.04522222280502319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,3072,0.025895110434956018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,2560,0.009683555199040307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,2048,0.008340444829728868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,3072,0.04307377669546339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,2560,0.0243857784403695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,2048,0.022825777530670166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,2560,0.0399582220448388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,1536,0.007204444044166141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,2048,0.03871644536654154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,1024,0.005843555761708154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,1536,0.01997599999109904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,65536,0.281351990169949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,1024,0.018294221825069852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,1536,0.03620355659061008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,768,0.0046817776229646476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,1024,0.032152000400755144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,768,0.01773244473669264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,512,0.0041866666740841335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,768,0.03164266546567281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,512,0.016861332787407767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,256,0.0037493333220481873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,512,0.030641777647866145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,128,0.003383999897374047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,256,0.015832889411184523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,256,0.028917332490285236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,128,0.01516888870133294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,64,0.0033475555893447665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,32,0.0034471110751231513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,64,0.01514400045077006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,65536,0.40229778819613987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,128,0.027407111393080816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,32,0.014911999305089315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,65536,0.1589182217915853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,16384,0.07304977708392672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,12288,0.05818844503826565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,16384,0.08252977662616305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,16384,0.041067557202445135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,12288,0.06967022021611531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,12288,0.033369776275422834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,10240,0.051826665798823036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,8192,0.044028444422615894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,10240,0.06434666448169284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,8192,0.05820444557401869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,7168,0.040417777167426214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,10240,0.029924445682101782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,8192,0.02511022157139248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,6144,0.018421333697107103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,7168,0.05550311009089152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,65536,0.267502228418986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,6144,0.03611822260750665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,7168,0.021734222769737244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,5120,0.015314666761292351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,6144,0.04850488901138306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,5120,0.03264888789918687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,4096,0.012230222423871359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,4096,0.029242667886945937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,5120,0.045175111956066556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,3584,0.012845333251688214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,3072,0.010547555155224271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,3584,0.027905778752432928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,4096,0.041088001595603094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,3584,0.040250665611690946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,3072,0.025794666674402025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,2560,0.008999110923873054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,3072,0.038498666551378034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,2048,0.00794666674402025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,2560,0.02423644396993849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,2048,0.0218542218208313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,2560,0.03635822070969476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,1536,0.0069004446268081665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,2048,0.03547022077772353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,1024,0.005203555441564984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,1536,0.01973777843846215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,1024,0.0181368887424469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,1536,0.033260444800059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,65536,0.3021226723988851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,768,0.004871110949251386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,1024,0.030887110365761652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,768,0.01737777723206414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,512,0.004111999852789773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,768,0.031153778235117596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,512,0.016674665941132438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,256,0.003644444462325838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,256,0.028375999795065984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,512,0.03014755580160353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,256,0.015595555305480957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,128,0.003354666754603386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,128,0.015079999963442484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,64,0.0031715554909573663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,32,0.014849777022997538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,32,0.0032559999575217566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,64,0.014866666661368476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,128,0.027218666341569688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,16384,0.07167377736833361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,16384,0.08250933223300509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,16384,0.03495555453830295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,12288,0.057890666855706104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,65536,0.1370799938837687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,12288,0.028176890479193792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,12288,0.06918133629692926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,10240,0.050867554214265615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,8192,0.0433555543422699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,10240,0.06394755840301514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,8192,0.056017776330312095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,7168,0.021173333128293354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,10240,0.025459556115998164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,7168,0.0402746664153205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,8192,0.02399377855989668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,7168,0.05338399940066867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,6144,0.018457776970333524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,6144,0.036095112562179565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,5120,0.017275555266274344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,6144,0.04785866538683573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,5120,0.0322746667597029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,4096,0.014663999279340109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,5120,0.04402044415473938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,4096,0.029095109966066148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,3584,0.012803555362754397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,4096,0.04044444362322489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,3584,0.027485332555241052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,3072,0.011830222275522021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,65536,0.2893039915296766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,65536,0.2669760121239556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,3584,0.039990223116344877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,3072,0.025739555557568867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,2560,0.00867733359336853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,3072,0.03821333249409994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,2560,0.023910222782029047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,2560,0.035819556978013777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,2048,0.00742133292886946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,2048,0.021328889661365088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,1536,0.006653333289755716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,2048,0.035868446032206215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,1536,0.019479999939600628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,1024,0.005147555636035072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,1536,0.03250311149491204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,1024,0.018018666240904067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,768,0.004512000001139111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,1024,0.030846221579445735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,768,0.01720533271630605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,512,0.003936000168323517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,768,0.03138844503296746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,512,0.016542222764756944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,256,0.0035777779089079965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,512,0.030054221550623577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,65536,0.10842400126987034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,256,0.015531554818153381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,128,0.0032995556377702286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,256,0.028472890456517536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,128,0.014953777194023132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,64,0.003132444495956103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,128,0.027129777603679236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,64,0.01484444406297472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,32,0.0032746667663256326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,32,0.014979556202888489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,16384,0.07270310984717475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,16384,0.07729511128531562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,12288,0.05755822526084053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,10240,0.02658044464058346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,12288,0.0647182199690077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,10240,0.0500204430686103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,12288,0.02354755500952403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,8192,0.021280889709790547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,16384,0.02847555610868666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,8192,0.043022221989101835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,10240,0.05913688739140829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,7168,0.01927911076280806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,8192,0.0528053343296051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,7168,0.03990133272276984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,6144,0.017031111650996737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,6144,0.0354417794280582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,7168,0.05086844497256809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,5120,0.015110222829712762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,6144,0.04619555672009786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,5120,0.03229511115286086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,4096,0.012512889173295764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,5120,0.042437334855397545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,3584,0.011423110961914062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,4096,0.02910311023394267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,65536,0.26734399795532227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,3584,0.027375110321574744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,3072,0.010287111004193624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,4096,0.03923733366860284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,2560,0.007818666597207388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,3072,0.024528889192475214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,3584,0.038719111018710665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,2560,0.0232622226079305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,3072,0.03676977753639221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,2048,0.006767999794748094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,2560,0.0336480008231269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,2048,0.02100177771515316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,1536,0.005528000079923206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,2048,0.03378044565518697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,1536,0.019545776976479422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,1024,0.018045332696702745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,1024,0.0046737777690092725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,65536,0.24763732486300996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,1024,0.030432889858881634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,1536,0.03200088938077291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,768,0.004278222305907143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,768,0.017145777742067974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,512,0.0038817777401871155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,768,0.030789332257376775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,256,0.0034595556143257353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,512,0.01645066671901279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,256,0.015459555718633862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,512,0.029726223813162908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,128,0.00328533330725299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,256,0.028427554501427546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,128,0.014912888407707214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,64,0.003060444361633725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,128,0.026888888743188646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,32,0.003168888803985384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,64,0.014828445182906257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,32,0.014744000302420722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,65536,0.08380266692903306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,16384,0.07225066423416138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,16384,0.09883289204703437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,12288,0.021172444025675457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,12288,0.056672890981038414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,10240,0.018106665876176622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,16384,0.021905778182877436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,10240,0.04948888884650337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,12288,0.08095288938946195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,8192,0.015507555670208402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,10240,0.07212533553441365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,8192,0.04286844531695048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,7168,0.014145778285132514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,8192,0.0636355545785692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,7168,0.03990844554371304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,6144,0.012679999901188744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,7168,0.0602782236205207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,5120,0.011136888629860349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,6144,0.03549955619706048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,5120,0.03205422229237027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,6144,0.05247999893294441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,4096,0.009655111365848118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,65536,0.26651912265353733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,5120,0.048968000544442065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,65536,0.3418586783938938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,4096,0.028808888461854722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,3584,0.009018667042255402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,4096,0.04445777667893303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,3072,0.008187555604510838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,3584,0.02623644471168518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,3072,0.024139554964171514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,3584,0.04262222184075249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,2560,0.00757244477669398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,3072,0.039099554220835366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,2560,0.02255200015174018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,2048,0.006406222366624408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,2560,0.03555466731389364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,2048,0.021240888370407954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,1536,0.005510222166776657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,2048,0.03505955470932855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,65536,0.0609991086853875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,1536,0.019318222999572754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,1024,0.004707555390066571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,1536,0.03417422374089559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,1024,0.01808533403608534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,768,0.004257777912749184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,768,0.017250667015711468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,1024,0.031268444326188825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,512,0.003863999827040566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,768,0.031176888280444678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,512,0.016332444217469957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,256,0.003481777591837777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,512,0.03011111087269253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,256,0.015528000063366361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,128,0.0031271111220121384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,256,0.02870133188035753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,128,0.014555555250909595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,64,0.002990222225586573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,32,0.003042666655447748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,64,0.015020444989204407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,128,0.02718044486310747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,32,0.014743111199802823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,16384,0.02217777735657162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,16384,0.07190933492448595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,12288,0.018608000543382432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,12288,0.05638133154975044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,16384,0.0961315565639072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,10240,0.016252444850073922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,10240,0.04965333475006951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,12288,0.07743555307388306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,8192,0.013616888887352414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,8192,0.04292533463901944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,10240,0.06805955701404147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,7168,0.012703999876976013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,7168,0.04014844364590115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,8192,0.061137775580088295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,6144,0.012039110892348819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,6144,0.03501688771777683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,7168,0.05856533182991875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,5120,0.010824888944625854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,6144,0.052460445298088916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,5120,0.030997332599427965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,4096,0.009289777941173976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,65536,0.2678497897254096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,5120,0.04767733481195238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,3584,0.008301333420806462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,3584,0.027186666925748188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,4096,0.02770400047302246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,65536,0.3037662241193983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,3584,0.039833777480655246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,4096,0.04209688968128628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,3072,0.00794400026400884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,65536,0.05134222242567274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,3072,0.023796444137891132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,2560,0.0069155556460221606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,3072,0.038351111941867404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,2560,0.022469333476490442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,2048,0.005807111246718301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,2560,0.035616000493367515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,2048,0.020961778031455144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,1536,0.005123555660247803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,2048,0.03484800126817491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,1536,0.019326221611764696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,1024,0.004412444515360726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,1536,0.03348177671432495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,1024,0.018160889546076458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,768,0.004024888906213972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,1024,0.030730668041441176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,768,0.017098666893111337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,512,0.0037413334680928122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,768,0.031255112753974065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,512,0.01635199950801002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,256,0.0033271111961868075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,512,0.030037333567937214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,256,0.015433778365453085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,128,0.0030666666312350165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,128,0.02699822187423706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,256,0.028875556257035997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,128,0.014762666490342883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,32,0.014639111028777229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,64,0.0029253332565228143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,32,0.0029786665820413162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,64,0.015050666199790107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,16384,0.0162453336848153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,16384,0.07193422317504883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,12288,0.014947555131382413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,12288,0.05563200182384915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,16384,0.07267022132873535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,10240,0.013623110949993134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,12288,0.05975377559661865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,10240,0.04871822065777249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,8192,0.012076444096035428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,10240,0.054939554797278516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,7168,0.011420444481902652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,8192,0.04278577698601616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,65536,0.26492977142333984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,8192,0.049550222025977246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,65536,0.22914044062296549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,7168,0.03882222374280294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,6144,0.010661332971519895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,65536,0.03949244485961066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,7168,0.04878222280078464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,6144,0.043863998519049756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,6144,0.03445155421892802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,5120,0.009939555492666032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,5120,0.0306693348619673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,4096,0.008963555925422246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,5120,0.04019822345839606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,4096,0.02721244427892897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,3584,0.008524444368150499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,4096,0.03594311078389486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,3584,0.025215110845035974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,3072,0.007826666865083907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,2560,0.022301332818137273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,3584,0.037385778294669256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,3072,0.02383822202682495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,2560,0.00639111093348927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,3072,0.03429955575201247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,2048,0.005648889061477449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,2560,0.03267822331852383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,2048,0.020636444290479023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,1536,0.005012444323963589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,2048,0.03221866819593642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,1536,0.019023999571800232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,1024,0.004288889053795072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,1536,0.03163999981350369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,768,0.003967111309369405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,1024,0.017457778255144756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,1024,0.03017866611480713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,768,0.01683555543422699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,512,0.029522667328516643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,512,0.003626666549179289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,768,0.030196444855795965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,512,0.016176889340082806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,256,0.0032924444725116095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,256,0.015365334020720588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,128,0.0030542222989930045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,128,0.014721777704026965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,256,0.02815999918513828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,64,0.0028986665937635633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,128,0.026611554953787062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,32,0.002918222298224767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,64,0.01462933255566491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,32,0.0147733340660731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,16384,0.011914666328165265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,16384,0.07120622528923883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,12288,0.010319111247857412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,65536,0.2681564490000407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,16384,0.07126310798856947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,10240,0.009268444445398118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,12288,0.05567200316323174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,65536,0.22886578241984049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,65536,0.032440887557135686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,12288,0.05958844555748833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,8192,0.00870488915178511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,10240,0.04907822277810839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,10240,0.05442755421002706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,8192,0.04080177678002252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,7168,0.008352888955010308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,8192,0.04946400059594048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,7168,0.03729688790109422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,6144,0.007343110938866933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,7168,0.047593779034084745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,6144,0.033462223079469465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,5120,0.008344888687133789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,6144,0.041747556792365186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,5120,0.03022222386466132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,4096,0.007638221813572778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,5120,0.03951377669970194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,4096,0.026828444666332666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,3584,0.0074480002125104266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,4096,0.035325331820382014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,3072,0.007141333487298753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,3584,0.0251866661840015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,3584,0.0351128876209259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,2560,0.006247111078765657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,3072,0.023617777559492324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,2560,0.02200800014866723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,3072,0.033810665210088096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,2048,0.005592888842026393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,2048,0.03171022070778741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,2560,0.03197155396143595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,1536,0.018813333577579923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,2048,0.020592888196309406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,1536,0.004893333133724001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,1024,0.0041982221106688184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,1024,0.01716888944307963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,768,0.0038924444880750445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,1536,0.031167109807332356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,768,0.016844444804721408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,1024,0.029917332861158583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,512,0.003567111161020067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,512,0.015871110889646742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,768,0.03001333276430766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,256,0.0032337777730491427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,512,0.029463112354278564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,256,0.015423110789722867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,128,0.003026666740576426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,256,0.02788177794880337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,128,0.014519999424616495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,64,0.0028862222615215513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,32,0.002875555513633622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,128,0.026750221848487854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,64,0.014521777629852295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,32,0.014730667074521383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,65536,0.02421422302722931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,16384,0.009310222334331935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,65536,0.2674249013264974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,16384,0.07053066624535455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,12288,0.008459555606047312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,65536,0.2278613249460856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,10240,0.007631999750932057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,16384,0.07090844710667928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,12288,0.055192000336117215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,12288,0.059360888269212514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,10240,0.04821422364976671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,8192,0.007601777712504069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,10240,0.054251554939481944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,7168,0.007232888705200619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,8192,0.04022044605678982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,7168,0.03704799877272712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,8192,0.04823288983768887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,6144,0.006887110984987683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,7168,0.045816888411839805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,6144,0.03370399938689338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,5120,0.007409777906205919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,6144,0.041173332267337374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,5120,0.02980177932315403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,4096,0.0069022224181228215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,5120,0.03813866774241129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,4096,0.026516444153255884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,3584,0.007345777418878343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,4096,0.03507733345031738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,3584,0.02524799936347538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,3072,0.007033777733643849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,3584,0.0350835555129581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,2560,0.006181333214044571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,3072,0.023584889041052923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,2560,0.021966222259733412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,3072,0.033738666110568576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,2048,0.00555288874440723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,2560,0.03200622068511115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,2048,0.020212444994184706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,1536,0.004846222284767362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,1024,0.017204445269372728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,2048,0.03218844532966614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,1536,0.01864622202184465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,1024,0.0041866666740841335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,1536,0.030905776553683813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,768,0.0037866665257347953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,1024,0.030047112041049536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,512,0.0035342222286595237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,512,0.029134223858515423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,768,0.0163102216190762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,768,0.029990222718980577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,512,0.015760888655980427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,256,0.0031866667171319327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,256,0.015154666370815702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,128,0.0029608888758553397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,128,0.014502222339312235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,256,0.028185778194003638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,64,0.002817777709828483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,32,0.002867555452717675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,64,0.014434667097197639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,32,0.014378666877746582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,65536,0.014017777310477363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,16384,0.0075431110130416016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,12288,0.007303111255168915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,16384,0.07018133004506429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,10240,0.00721066693464915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,65536,0.2667466534508599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,12288,0.05559555689493815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,8192,0.007171555525726742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,10240,0.0465848876370324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,7168,0.00702044450574451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,8192,0.040334221389558576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,6144,0.006823110911581252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,7168,0.036963555547926165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,5120,0.007223111059930589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,6144,0.03318311108483209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,4096,0.0068764446510208984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,5120,0.029659556017981634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,4096,0.02663555575741662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,3584,0.007280888656775157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,3072,0.006952889263629913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,3584,0.025063110722435847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,2560,0.006144888699054718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,3072,0.023558222585254248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,2048,0.005515555540720622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,2560,0.02184533410602146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,1536,0.0047937776479456155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,2048,0.020225778222084045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,1024,0.0041866666740841335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,1536,0.018546667363908555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,768,0.0037866665257347953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,256,0.003105777833196852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,1024,0.01719111038578881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,512,0.0034871109657817413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,768,0.01646933290693495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,512,0.01585422290696038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,32,0.0028000000036425064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,128,0.0029333333174387612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,256,0.014939554863505893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,64,0.0028737777223189673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,128,0.014478221535682678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,64,0.014463999205165438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,65536,0.01368533323208491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,32,0.014483556151390076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,16384,0.007159111400445302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,12288,0.007459555235173967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,16384,0.0699217783080207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,10240,0.007243555453088548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,65536,0.2688808970981174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,12288,0.05487199955516391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,8192,0.007167999943097432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,10240,0.04649777876006233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,7168,0.00699733363257514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,8192,0.03993689020474752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,6144,0.006738666858938005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,7168,0.03689688775274489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,5120,0.0070977773931291364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,6144,0.03323910964859857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,4096,0.006807111203670502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,5120,0.030062221818500098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,3584,0.007122666471534305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,4096,0.026544888814290363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,3072,0.006776888751321369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,3584,0.025077333052953083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,2560,0.006150222072998683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,3072,0.023472888602150813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,2048,0.005491555564933353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,2560,0.02190133266978794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,1536,0.00481599983241823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,2048,0.020097777247428894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,1024,0.004152889053026835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,1536,0.018583110637134977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,768,0.0037919998996787598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,1024,0.017051556044154696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,512,0.0034808889031410217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,768,0.01646311084429423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,256,0.00314666661951277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,512,0.015608888533380298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,128,0.002934222213096089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,64,0.014516444669829475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,256,0.015006222658687167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,64,0.0027831111931138565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,128,0.014474666780895658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,32,0.002817777709828483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,32,0.014308444327778287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,12288,0.32039377424452037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,16384,0.4148506588406033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,10240,0.48924530877007383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,12288,0.5827920171949599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,10240,0.27080000771416557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,128,0.02684088879161411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,8192,0.21779021951887342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,8192,0.39319645033942324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,16384,0.7671706411573621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,7168,0.20077866978115508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,6144,0.16855288876427543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,7168,0.34332977400885684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,6144,0.2987111144595676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,12288,0.3759804566701253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,16384,0.4493458006117079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,5120,0.15967733330196804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,5120,0.2497404416402181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,4096,0.11765600575341119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,4096,0.20281155904134116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,3584,0.10760888788435195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,3584,0.177964448928833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,10240,0.31984443134731716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,8192,0.26170489523145885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,3072,0.09518933296203613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,3072,0.1546728875901964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,2560,0.13046577241685656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,2560,0.09001955721113418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,7168,0.23258490032619902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,2048,0.06957244210773043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,2048,0.10662133163876003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,1536,0.08286666870117188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,1536,0.057180444399515785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,6144,0.20178043842315674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,1024,0.058509336577521436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,1024,0.04946488804287381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,5120,0.17441956202189127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,768,0.046634667449527316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,768,0.03972622089915805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,4096,0.15074933899773493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,512,0.03412533468670315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,512,0.03365244468053182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,3584,0.128257777955797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,256,0.021388444635603163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,256,0.026745777991082933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,3072,0.117777771419949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,128,0.011969777444998423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,128,0.02230666743384467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,2560,0.10488977697160508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,64,0.012791111237472959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,64,0.025044444534513686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,32,0.013704000247849358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,32,0.025723555021815833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,2048,0.08811022175682916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,256,0.04253066579500834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,1536,0.07671199904547797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,512,0.05139466789033678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,128,0.03706311186154684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,768,0.05785955323113335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,1024,0.06274488899442884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,16384,0.13006933530171713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,12288,0.10349866416719224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,16384,0.19158311684926352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,65536,0.46943198310004336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,12288,0.14792088667551676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,10240,0.08955022361543442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,10240,0.12275821632809109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,8192,0.07284266418880887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,8192,0.09949155648549397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,7168,0.08723644415537517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,7168,0.06587821907467313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,6144,0.07698222001393636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,6144,0.059078223175472684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,65536,0.7852177619934082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,16384,0.150489780637953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,5120,0.05154666635725233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,5120,0.06465688678953382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,12288,0.1223564412858751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,4096,0.043976889716254346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,4096,0.052998221582836576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,10240,0.11716000239054362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,3584,0.04066310988532172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,3584,0.0469742218653361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,8192,0.09248711003197564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,65536,0.5266550911797417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,3072,0.0470382206969791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,3072,0.037088887559043035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,2560,0.03377866744995117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,2560,0.03544888893763224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,2048,0.029522667328516643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,2048,0.030947556098302204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,7168,0.08473155233595107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,1536,0.023598222268952265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,1536,0.02611733310752445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,6144,0.07530755466885038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,1024,0.016559999850061204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,1024,0.02288355595535702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,5120,0.06761866807937622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,768,0.012911111116409302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,768,0.020814221766259935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,4096,0.060565332571665444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,512,0.009653333160612319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,3584,0.05806399716271294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,512,0.019480889042218525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,256,0.006911110960774952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,256,0.016151999433835346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,3072,0.0539591113726298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,128,0.0048142220411035745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,128,0.015223999818166098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,2560,0.05089333322313097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,64,0.0042444442709287005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,64,0.015760888655980427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,32,0.004569777597983678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,32,0.01570222278436025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,2048,0.0469777782758077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,1536,0.043282667795817055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,1024,0.038373334540261164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,768,0.03744533326890733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,512,0.03599111239115397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,16384,0.1121706697675917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,256,0.03067200051413642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,128,0.02794044547610813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,16384,0.14263821972741023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,65536,0.4024542172749837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,12288,0.08960800038443671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,12288,0.10887110895580715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,10240,0.0848346683714125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,10240,0.09258755710389878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,8192,0.06334133280648126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,8192,0.07412622372309367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,65536,0.583263980017768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,7168,0.057809776730007596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,7168,0.06526844369040595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,6144,0.05750844213697645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,6144,0.050965332322650485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,5120,0.048657778236601085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,5120,0.04494222336345249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,16384,0.13438310888078478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,12288,0.10715644227133857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,4096,0.03881866733233134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,4096,0.039975110027525164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,3584,0.036736889017952815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,10240,0.09532977475060357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,65536,0.4788835313585069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,3584,0.036200887627071805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,3072,0.032906668053732976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,3072,0.031543999910354614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,8192,0.08021333482530382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,2560,0.03013422091801961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,2560,0.027320888307359483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,7168,0.07917600207858615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,2048,0.02749600013097127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,2048,0.02260533306333754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,1536,0.01815822223822276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,1536,0.02405422263675266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,6144,0.06696266598171659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,1024,0.01259022206068039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,1024,0.021172444025675457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,5120,0.0612000028292338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,768,0.010298666854699453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,768,0.02020355562369029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,3584,0.05321777860323588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,512,0.007996444072988298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,512,0.018447111050287884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,4096,0.05522400140762329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,256,0.005412444472312927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,256,0.015962666935390897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,2560,0.047319110896852284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,128,0.004279110994603899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,128,0.015440000428093804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,768,0.03654311100641886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,3072,0.04923289020856222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,64,0.00396799999806616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,32,0.004324444466167026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,64,0.015437333120240105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,32,0.01575911045074463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,1536,0.04033777780003018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,2048,0.04187466700871786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,1024,0.03679555654525757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,16384,0.12044266859690349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,16384,0.10448622041278416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,12288,0.0928186641799079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,65536,0.3742382261488173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,12288,0.08334755235248142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,128,0.027686221732033625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,10240,0.07812355624304877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,256,0.02950577934583028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,10240,0.07219911283916898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,8192,0.059262222713894315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,8192,0.06294311417473687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,65536,0.49765777587890625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,7168,0.05814311239454481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,512,0.034865776697794594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,7168,0.0557617810037401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,6144,0.04858577913708157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,6144,0.049490667051739164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,5120,0.041724443435668945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,5120,0.04290666513972812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,16384,0.12080978022681342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,4096,0.03417955504523383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,4096,0.0374257763226827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,65536,0.4281920062171088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,12288,0.10520178079605103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,3584,0.030672887961069744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,10240,0.08902044428719415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,3584,0.03463022245301141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,3072,0.031697776582505964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,3072,0.027127110295825537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,2560,0.023316444622145757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,2560,0.02860355708334181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,8192,0.07548889186647204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,2048,0.01944177846113841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,7168,0.0705359975496928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,2048,0.026551999979548987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,1536,0.015113777584499784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,1536,0.02279022170437707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,6144,0.06399466594060262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,1024,0.010828444527255164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,1024,0.021359110871950786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,5120,0.05845955345365736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,768,0.009142222503821055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,768,0.019688889384269714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,4096,0.0524871117538876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,512,0.00721599989467197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,512,0.01677688956260681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,3584,0.05065600077311198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,256,0.004784000002675586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,256,0.015814221567577787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,2560,0.045137776268853076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,128,0.004089777668317159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,128,0.015098666151364645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,3072,0.04738666613896688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,64,0.003765333443880081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,64,0.0155395550860299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,32,0.003949333396222857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,32,0.015504888362354703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,768,0.03581510980923971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,1536,0.03913244605064392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,2048,0.04001155495643616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,1024,0.03592977921168009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,16384,0.10933955510457356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,16384,0.08399022287792629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,65536,0.28646310170491535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,12288,0.07463022073109944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,12288,0.06857066684299044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,10240,0.06339999702241686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,10240,0.06003555324342516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,128,0.027034666803148057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,256,0.028800888193978205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,8192,0.048841777775022716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,65536,0.40943111313713926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,8192,0.05046933227115207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,7168,0.04486844605869717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,7168,0.05235644512706333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,512,0.03359999921586778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,6144,0.04007200068897671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,6144,0.04059733284844293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,5120,0.03389422098795573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,5120,0.03633688886960348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,16384,0.10619378089904785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,4096,0.02814488940768772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,4096,0.031727999448776245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,65536,0.3561031023661296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,12288,0.08788088957468669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,3584,0.02962133288383484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,3584,0.025028443998760645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,3072,0.022314666046036616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,3072,0.02745955520206028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,10240,0.07844799757003784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,2560,0.019101333287027147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,2560,0.025015999873479206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,8192,0.06714222166273329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,2048,0.015829333000712924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,2048,0.023767110374238756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,7168,0.06234133243560791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,1536,0.012248000337017907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,1536,0.020998222960366145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,6144,0.05739021963543362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,1024,0.0092604441775216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,1024,0.019504888190163504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,5120,0.05227555500136482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,768,0.007852444218264686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,768,0.017603556315104168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,3584,0.046316444873809814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,4096,0.04796977837880453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,512,0.006562666760550604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,512,0.01571822166442871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,256,0.004118222329351637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,256,0.015065777632925244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,3072,0.0429422226217058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,128,0.0037768888804647657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,128,0.014712888333532544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,2560,0.04113511244455973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,64,0.0034506666577524612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,64,0.014614222778214348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,32,0.0037946667936113146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,32,0.014704888065656027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,1536,0.03732622332043118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,2048,0.03904622130923801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,1024,0.03481511274973551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,768,0.035079111655553184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,65536,0.26509067747328013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,16384,0.07813422547446357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,16384,0.09584089120229085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,256,0.028239998552534316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,512,0.03232889042960273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,128,0.02718399961789449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,12288,0.06296355856789483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,12288,0.07567022244135539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,10240,0.05633066760169136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,10240,0.05608266592025757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,8192,0.0462773342927297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,8192,0.04528888728883532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,7168,0.04261599977811178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,7168,0.04078933265474107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,65536,0.36440443992614746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,6144,0.036372443040212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,6144,0.038740442858801946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,5120,0.030453334252039593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,5120,0.0344177782535553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,16384,0.1015617781215244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,65536,0.3259982268015544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,4096,0.02472177810139126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,4096,0.029996444781621296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,3584,0.022181333767043218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,3584,0.027927110592524212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,12288,0.08372177680333455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,3072,0.019682667321628995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,10240,0.07515555620193481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,3072,0.025974222355418738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,2560,0.016744888491100736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,2560,0.023786667320463393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,8192,0.0643635524643792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,2048,0.013431111143694984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,2048,0.022069333328141108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,7168,0.06012355619006687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,1536,0.010986666712496014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,1536,0.020463999774720933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,6144,0.057484441333346896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,1024,0.008575999902354347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,1024,0.01937955617904663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,5120,0.050607111718919545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,768,0.007298666569921706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,768,0.01754755609565311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,4096,0.045681777927610606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,512,0.005787555542257097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,512,0.015564444992277356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,3584,0.044213334719340004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,256,0.0038977778620190094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,256,0.015044444137149386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,3072,0.04144444399409824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,128,0.003510222252872255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,128,0.014310222533014087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,2560,0.0403404434521993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,64,0.0032435556252797446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,64,0.014510222607188754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,32,0.003591999825504091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,32,0.014519110321998596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,2048,0.037646220790015325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,1536,0.03734488950835334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,1024,0.03523022267553542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,768,0.03495999839570787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,65536,0.24512354532877603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,16384,0.08289066950480144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,16384,0.07288266552819146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,12288,0.05967289209365845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,12288,0.057321780257754855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,65536,0.32354044914245605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,10240,0.0485831101735433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,10240,0.05667199691136678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,512,0.030604445272021826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,256,0.02832177612516615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,8192,0.039260443713929914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,8192,0.04320444332228767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,128,0.02679466704527537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,7168,0.03978311022122701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,7168,0.03474222289191352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,6144,0.03095910946528117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,6144,0.036292443672815956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,5120,0.02637422250376807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,5120,0.03216177887386746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,16384,0.1027688913875156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,4096,0.021914665897687275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,4096,0.028512001037597656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,65536,0.3166666560702854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,12288,0.0849368903372023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,3584,0.02640177806218465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,3584,0.01958044370015462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,3072,0.01958755486541324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,3072,0.024932444095611572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,10240,0.07707822322845459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,2560,0.014679999815093147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,2560,0.02305511136849721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,8192,0.06505599949094984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,2048,0.011957333319716983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,2048,0.021371554997232225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,7168,0.060602664947509766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,1536,0.009924444059530893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,1536,0.01984177695380317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,6144,0.05570133527119955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,1024,0.007755555212497711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,1024,0.018048000004556444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,5120,0.050684445434146456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,768,0.006703111032644908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,768,0.016324443949593436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,4096,0.04506311151716444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,512,0.005010666532648934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,512,0.015677332878112793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,3584,0.04380888740221659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,256,0.003839111162556542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,256,0.015185778339703878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,3072,0.04116444455252753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,128,0.0034328889515664843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,128,0.01443555619981554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,2560,0.040417777167426214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,64,0.0032213332338465583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,64,0.014701333310869006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,32,0.0036159998012913596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,32,0.01458666721979777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,1536,0.03608444333076477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,2048,0.037774221764670476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,768,0.033194667763180204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,1024,0.03371466530693902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,16384,0.06786221928066678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,16384,0.07055377960205078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,65536,0.2274657885233561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,12288,0.05532177620463901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,12288,0.05628533495797051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,10240,0.048792888720830284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,10240,0.04921244581540426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,65536,0.26558044221666127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,8192,0.03896266553137038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,8192,0.040525333748923406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,128,0.026838221483760413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,256,0.02836622132195367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,512,0.03003022074699402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,7168,0.03758488761054145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,7168,0.029934220843844946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,6144,0.03387644555833604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,6144,0.027110222313139174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,5120,0.022760000493791368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,5120,0.030284444491068523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,16384,0.08901333146625096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,4096,0.018815999229749043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,4096,0.026891556051042344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,65536,0.26421599917941624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,12288,0.07392799854278564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,3584,0.02498488956027561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,3584,0.01664355562792884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,3072,0.014569777581426831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,3072,0.023584889041052923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,10240,0.06709066364500257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,2560,0.01253600004646513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,2560,0.022227555513381958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,8192,0.05817511346605089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,2048,0.01055555542310079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,2048,0.020870221985710993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,6144,0.04964355627695719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,1536,0.008805333740181392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,7168,0.054359998967912465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,1536,0.019679110911157396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,1024,0.00701155596309238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,1024,0.01774488886197408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,5120,0.045736889044443764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,768,0.006234666539563074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,768,0.01626399987273746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,4096,0.04212622178925408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,512,0.00425511101881663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,3584,0.040660444233152605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,512,0.015744000673294067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,256,0.0038151111867692736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,256,0.015116444892353482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,3072,0.03842488924662272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,128,0.0033528889632887314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,128,0.014295111099878946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,2560,0.038034667571385704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,64,0.0031555555760860443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,64,0.014360000689824423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,32,0.0034666665726237824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,32,0.014355555176734924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,2048,0.0363031095928616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,1536,0.035813334915373064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,1024,0.033640000555250384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,768,0.03128088845147027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,16384,0.06424178017510308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,16384,0.06023199690712822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,65536,0.20506844255659315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,65536,0.20756178432040742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,12288,0.053620444403754346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,12288,0.04013155566321479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,10240,0.034882667991850115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,10240,0.046910223033693105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,8192,0.027951998843087092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,256,0.02810488806830512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,512,0.02964622113439772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,8192,0.03881422347492642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,128,0.02641333308484819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,7168,0.03562044435077243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,7168,0.025034666061401367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,6144,0.03204088740878635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,6144,0.022641777992248535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,5120,0.019293333093325298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,5120,0.029120888974931505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,65536,0.23944534195793998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,16384,0.0822497804959615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,4096,0.025981333520677354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,4096,0.015778667396969266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,12288,0.06909866465462579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,3584,0.014064888159434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,3584,0.02390577726893955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,3072,0.012312889099121094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,3072,0.022758222288555566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,10240,0.06244444184833103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,2560,0.010832889212502373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,2560,0.021916444102923077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,8192,0.054526223076714404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,2048,0.009390222529570261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,7168,0.05093155635727776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,2048,0.02075111038155026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,1536,0.007989333735571967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,1536,0.018603555030292936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,6144,0.046689778566360474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,1024,0.006435555716355641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,1024,0.016351110405392118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,5120,0.04346133271853129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,768,0.0053093334039052325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,768,0.01607111096382141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,4096,0.03981155488226149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,3584,0.039389332135518394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,512,0.00434933313065105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,512,0.015602666470739575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,256,0.003779555360476176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,3072,0.03771911064783732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,256,0.015006222658687167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,128,0.0033973333322339584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,128,0.014362666341993542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,2560,0.037054220835367836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,64,0.0032479998966058097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,64,0.014289778139856128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,32,0.0035502223504914176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,32,0.01423288881778717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,1536,0.03487022386656867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,2048,0.0349333319399092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,1024,0.02978044417169359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,16384,0.053512000375323825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,16384,0.06241866615083483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,65536,0.2070773310131497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,768,0.030930668115615845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,12288,0.03634044528007507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,65536,0.1799795495139228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,12288,0.050399998823801674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,10240,0.045477333996031016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,10240,0.03062044580777486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,8192,0.02502311103873783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,8192,0.037104000647862755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,256,0.02810488806830512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,512,0.03013422091801961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,128,0.026337777574857075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,7168,0.03455466694302029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,7168,0.022728888524903193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,6144,0.030260443687438965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,6144,0.019920888874265883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,5120,0.017343110508388944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,5120,0.028035554620954726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,16384,0.07876622014575534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,4096,0.013735999663670858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,4096,0.024874667326609295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,65536,0.23169332080417207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,3584,0.012539555629094442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,3584,0.023754666248957317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,12288,0.06689155764049955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,3072,0.011640888949235281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,3072,0.02256977723704444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,10240,0.06089599927266439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,2560,0.010297777752081553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,2560,0.021494223011864558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,8192,0.05268177721235487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,2048,0.008829333715968663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,2048,0.020100444555282593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,7168,0.04941066768434313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,1536,0.007717333734035492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,1536,0.0177839994430542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,6144,0.04538933436075846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,1024,0.006121777825885349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,1024,0.016515556308958266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,5120,0.04251111216015286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,768,0.004975999808973736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,768,0.016006223029560514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,4096,0.03891466723548041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,512,0.0042444442709287005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,512,0.01533777846230401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,3584,0.03879022267129686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,256,0.003657777690225177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,256,0.014750222365061441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,3072,0.03679200013478597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,128,0.003474666840500302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,128,0.014268444644080268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,2560,0.03700000047683716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,64,0.003294222263826264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,64,0.014374223020341663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,32,0.0033200000309281876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,2048,0.03453155689769321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,32,0.01425244410832723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,1536,0.03460444344414605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,1024,0.029993779129452173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,65536,0.20447911156548393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,768,0.03042666779624091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,16384,0.060460441642337374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,65536,0.15167821778191462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,16384,0.04452355702718099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,12288,0.04987022280693054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,12288,0.03605066736539205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,10240,0.04349422123697069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,10240,0.03164622187614441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,8192,0.022775999373859827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,8192,0.03567644291453891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,512,0.029168887270821467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,256,0.027942223681343928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,7168,0.020261334048377145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,7168,0.03293777836693658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,6144,0.01796888808409373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,6144,0.03018666638268365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,128,0.026483555634816486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,5120,0.01643733349111345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,5120,0.027250667413075764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,16384,0.07678755786683825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,4096,0.012856889102194043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,12288,0.06487466891606648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,4096,0.02488800055450863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,3584,0.011422221859296164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,65536,0.2225137816535102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,3584,0.023634667197863262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,3072,0.01071733319097095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,10240,0.05844533443450928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,3072,0.0225911107328203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,2560,0.009666666388511658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,2560,0.021359110871950786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,8192,0.05110133356518216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,2048,0.00835111074977451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,2048,0.01981599960062239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,7168,0.047614223427242704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,1536,0.007277333074145847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,1536,0.017286222842004564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,6144,0.04366044534577263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,1024,0.006036444256703059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,1024,0.016149333781666227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,5120,0.04169777697987027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,768,0.004871110949251386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,768,0.015722667177518208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,3584,0.03828533490498861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,4096,0.039055999782350324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,512,0.004247111164861255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,512,0.015336889359686108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,256,0.003669333126809862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,256,0.014852444330851236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,3072,0.03609955641958449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,128,0.0033128888656695685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,128,0.014223999447292753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,2560,0.03671733207172818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,64,0.0031600000543726813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,64,0.014301333162519666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,32,0.0032906666811969546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,32,0.014244443840450711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,2048,0.03452444407675002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,1536,0.033942222595214844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,1024,0.029834667841593426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,65536,0.19395821624332002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,16384,0.0581173300743103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,768,0.030605332718955144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,16384,0.039346665143966675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,12288,0.04658755660057068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,65536,0.13193955686357287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,12288,0.02906044324239095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,10240,0.04123199979464213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,10240,0.024811555941899616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,8192,0.02047644390000237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,8192,0.03377688924471537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,512,0.029492444462246362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,256,0.027963555521435205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,7168,0.031183110343085393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,128,0.026200888885392085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,7168,0.01844800015290578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,6144,0.016223111086421542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,6144,0.028954668177498713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,5120,0.01350844403107961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,5120,0.026346666945351496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,16384,0.074617776605818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,12288,0.0630906687842475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,65536,0.21738755702972412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,4096,0.011506666739781698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,4096,0.023960888385772705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,3584,0.010597333312034607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,3584,0.0229795558585061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,3072,0.009702222214804756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,10240,0.05752711163626777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,3072,0.02168088820245531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,2560,0.00870577742656072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,2560,0.020502222908867728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,8192,0.04933244321081373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,2048,0.007724444071451823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,2048,0.019357333580652874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,6144,0.042760888735453285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,7168,0.04675199919276767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,1536,0.006689777804745569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,1536,0.01687733332316081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,5120,0.04065333472357856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,1024,0.004969777746333016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,1024,0.016102222932709586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,768,0.004394666602214177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,768,0.01552888916598426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,3584,0.038183109627829656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,4096,0.037703110112084284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,512,0.003934222377008862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,512,0.015096889601813423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,256,0.003544888976547453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,256,0.014554666148291694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,3072,0.03608355588383145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,128,0.0032595555401510666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,128,0.014149333039919535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,2560,0.03608622153600057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,64,0.003108444520168834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,64,0.01408622165520986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,32,0.0031431110368834603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,32,0.013950222068362765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,2048,0.034183998902638756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,65536,0.18979556030697295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,65536,0.10685511430104573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,16384,0.03353955679469638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,16384,0.05563644568125407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,1536,0.03288266725010342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,1024,0.029341333442264136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,12288,0.044329779015647046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,12288,0.026311111119058397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,10240,0.023103111320071753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,10240,0.03867288761668735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,768,0.02990844514634874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,8192,0.01898755629857381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,8192,0.03293777836693658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,512,0.02940355406867133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,7168,0.017328888177871704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,7168,0.030477331744299993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,256,0.0276826669772466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,6144,0.01581333412064446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,128,0.02642755541536543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,6144,0.02819733487235175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,5120,0.013656888571050433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,5120,0.025997334056430395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,65536,0.20797155963049996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,4096,0.01165600038237042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,4096,0.02405333353413476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,16384,0.07223466369840834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,3584,0.010183999935785929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,3584,0.022965333527988855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,12288,0.060292442639668785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,3072,0.009108444054921469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,3072,0.021083555287784998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,10240,0.0551128884156545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,2560,0.007970666719807519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,2560,0.019110222657521565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,8192,0.04763377706209818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,2048,0.007151111132568783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,2048,0.017857778403494094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,7168,0.04502222273084852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,1536,0.005924444645643234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,1536,0.016904888881577384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,6144,0.04198399848408169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,1024,0.004950222041871813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,1024,0.01588800052801768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,5120,0.039910223748948835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,768,0.004345777961942884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,768,0.015418666932317944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,4096,0.03713688916630215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,512,0.0038453332251972626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,512,0.015358222855461968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,3584,0.03718577822049459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,256,0.0036675557494163513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,256,0.014431110686726041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,3072,0.03533066643608941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,128,0.003133333391613431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,128,0.014192889134089151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,2560,0.03454666667514377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,64,0.002970666728085942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,64,0.01425155500570933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,32,0.003043555551105075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,32,0.014165333575672574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,2048,0.03288799855444167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,65536,0.18077422512902153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,65536,0.0844808883137173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,16384,0.026328000757429335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,16384,0.053039110369152494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,1536,0.031368000639809504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,1024,0.029599110285441082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,12288,0.04275022281540764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,768,0.03000266684426202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,12288,0.021180444293551978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,10240,0.018509333332379658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,10240,0.037145776881111994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,8192,0.015673778123325773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,8192,0.03137333194414774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,512,0.029422223567962646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,7168,0.014312888185183207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,256,0.027609777119424608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,7168,0.029831111431121826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,128,0.02608888844648997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,6144,0.012602667013804117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,6144,0.027436445156733196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,5120,0.011224889092975192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,5120,0.02525955604182349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,65536,0.2026453283098009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,4096,0.009648000200589498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,4096,0.022877333892716303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,16384,0.06959555546442668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,3584,0.009034666750166152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,3584,0.021819555097156104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,12288,0.058627552456325956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,3072,0.008378666308191087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,3072,0.020181333025296528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,10240,0.05312444435225593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,2560,0.007733333441946242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,2560,0.01848355597919888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,8192,0.0466177761554718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,2048,0.006593777901596493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,2048,0.017810665898852877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,6144,0.04107111030154758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,7168,0.04462933209207323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,1536,0.00563733321097162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,1536,0.016785777277416654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,1024,0.004693333473470476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,1024,0.015973332855436537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,5120,0.03935911258061727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,768,0.004217777815130022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,768,0.015446222490734525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,4096,0.03666399916013082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,512,0.0038044444388813446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,512,0.014985778265529208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,3584,0.03718044360478719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,256,0.0033822223130199644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,256,0.014708444476127625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,3072,0.035190221336152815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,128,0.0031315556002987754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,128,0.014083556003040738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,2560,0.032842665910720825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,64,0.0029537777105967202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,64,0.014035556051466199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,32,0.003000000077817175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,32,0.013882666826248169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,2048,0.031297776434156634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,65536,0.17940178182390001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,65536,0.059318224589029946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,16384,0.019698666201697457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,16384,0.04988177617390951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,1536,0.03120888935195075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,12288,0.015303111738628812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,1024,0.02903199858135647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,12288,0.04047377904256185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,10240,0.013430222041077085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,10240,0.03615199857287937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,768,0.030004445049497817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,8192,0.011900444825490316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,8192,0.031199110878838435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,512,0.02905333373281691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,7168,0.012684444586435953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,7168,0.028804444604449805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,256,0.027865777413050335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,6144,0.010849778023031024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,6144,0.02690222197108799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,128,0.02628533376587762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,5120,0.010709332923094431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,5120,0.02492888934082455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,16384,0.06641688611772326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,4096,0.00960977789428499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,65536,0.1986560026804606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,4096,0.021735999319288466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,3584,0.008377778033415476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,3584,0.020409777760505676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,12288,0.05594755543602837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,3072,0.007876444194051955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,3072,0.019481778144836426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,10240,0.051608890295028687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,2560,0.00720355576939053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,2560,0.018495111001862418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,8192,0.04563466707865397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,2048,0.006028444402747684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,2048,0.017459554804695975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,7168,0.04323822259902954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,1536,0.0052435555391841466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,1536,0.016540444559521146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,6144,0.04112977782885233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,1024,0.00434400017062823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,1024,0.015956444872750174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,5120,0.03901422354910109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,768,0.004090666770935059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,768,0.01535022258758545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,4096,0.03635288940535651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,512,0.003693333516518275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,512,0.015040000279744467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,3584,0.03461866577466329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,256,0.0033315554675128725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,256,0.014431110686726041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,3072,0.03311911225318909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,128,0.003101333354910215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,128,0.014226666755146451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,2560,0.03253688746028476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,2048,0.030952887402640447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,64,0.0029813332690132987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,32,0.0030586665703190696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,64,0.014045332868893942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,32,0.01405244403415256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,65536,0.17735111713409424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,65536,0.054692443874147206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,16384,0.01573955516020457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,1536,0.030819555123647053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,16384,0.04924266537030538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,12288,0.013520888984203339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,12288,0.0399911105632782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,1024,0.02939288814862569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,10240,0.012615111139085559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,10240,0.03595022360483805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,768,0.03016888764169481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,8192,0.011205332974592844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,8192,0.030844443374209937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,512,0.029072887367672388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,7168,0.010671111444632212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,256,0.027315555347336665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,7168,0.028817776176664565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,6144,0.010107555323176915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,6144,0.026198221577538386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,128,0.025991110338105097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,5120,0.010017777482668558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,5120,0.024015110399987962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,65536,0.19878933164808485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,4096,0.008998221821255153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,4096,0.021716444028748408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,16384,0.06562311119503446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,3584,0.008397333323955536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,12288,0.055177778005599976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,3584,0.020527111159430612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,3072,0.007672000262472365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,10240,0.051627556482950844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,3072,0.019344889455371432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,2560,0.00638488887084855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,2560,0.018364444375038147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,8192,0.04561333192719353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,2048,0.005658666706747479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,2048,0.017489777670966256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,7168,0.043040888177024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,1536,0.004962666581074397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,1536,0.016520000166363187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,6144,0.04051822092798021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,1024,0.004274666723277834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,1024,0.015799111790127225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,5120,0.03642044464747111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,768,0.003907555507289039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,768,0.015391111373901367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,4096,0.03550844391187032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,512,0.003681777666012446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,512,0.015145777000321282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,3584,0.033929778469933405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,256,0.0032497776879204642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,256,0.014484445254007975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,3072,0.03271289004219903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,128,0.003023111157947116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,128,0.01424800025092231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,2560,0.032307555278142296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,64,0.002875555513633622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,64,0.014090667168299357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,32,0.0029057777590221832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,32,0.013850666582584381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,2048,0.030431999100579157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,65536,0.04201244314511617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,65536,0.1747431092792087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,16384,0.012669333153300814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,16384,0.04828977915975782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,1536,0.030664887693193223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,12288,0.011131555669837527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,12288,0.03933599922392104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,1024,0.029458665185504492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,10240,0.010501333408885531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,10240,0.03564088874393039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,768,0.029687109920713637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,8192,0.009536888864305284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,8192,0.03023911184734768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,512,0.029299557209014893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,7168,0.00909511082702213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,256,0.027553778555658128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,7168,0.028176890479193792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,128,0.02644088864326477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,6144,0.010454222559928894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,6144,0.025703110628657874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,5120,0.00959022260374493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,5120,0.023342221975326538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,16384,0.06491111384497748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,65536,0.1977377732594808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,4096,0.008264000217119852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,4096,0.021399999658266704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,3584,0.007520000139872233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,3584,0.020946666598320007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,12288,0.054400000307295054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,3072,0.00723555518521203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,3072,0.01925066610177358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,10240,0.051422221793068774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,2560,0.0063688887490166565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,2560,0.01847199930085076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,8192,0.0451822214656406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,2048,0.005670222143332164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,2048,0.017510222064124215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,7168,0.04314133193757799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,6144,0.0399528874291314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,1536,0.004951999833186467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,5120,0.03791466686460707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,1536,0.01658933361371358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,1024,0.004266666869322459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,1024,0.015785778562227886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,768,0.003955555458863576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,768,0.01534222231970893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,4096,0.033999111917283796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,512,0.003590222034189436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,512,0.014889778362380134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,3584,0.03386133246951633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,256,0.0032248888164758682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,256,0.014193778236707052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,3072,0.03198844525549147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,128,0.0029973333908451926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,128,0.01388533330625958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,2560,0.03232889042960273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,64,0.0029582221888833572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,64,0.01406133340464698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,32,0.0031866667171319327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,32,0.014004444082578024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,65536,0.029141333368089464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,65536,0.17182311746809217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,2048,0.030559112628300984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,16384,0.01091822236776352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,16384,0.047672000196244985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,1536,0.030906667311986286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,12288,0.009704888694816166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,12288,0.03864444295565287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,1024,0.02920266654756334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,10240,0.008619555168681675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,768,0.02976177798377143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,10240,0.03506222367286682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,512,0.028912001185946997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,8192,0.007801777786678738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,8192,0.029751999510659113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,7168,0.007530666887760162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,7168,0.027099554737408955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,256,0.027293332748942908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,6144,0.006984000404675801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,6144,0.025039111574490864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,128,0.02602844436963399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,5120,0.007415999968846639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,5120,0.02318222158484989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,65536,0.19747289021809897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,4096,0.0070293330483966405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,4096,0.021236444513003033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,16384,0.06398666567272611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,3584,0.0073626670572492815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,3584,0.020246222615242004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,12288,0.05439466569158766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,3072,0.0070568886068132185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,3072,0.01921422282854716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,10240,0.05094666613472832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,8192,0.043643554051717125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,2560,0.006266666783226862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,2560,0.018336888816621568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,2048,0.005562666803598404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,7168,0.04173688756095039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,2048,0.017453332742055256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,1536,0.00489688871635331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,6144,0.03828622235192193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,1536,0.01660711069901784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,1024,0.0042071110672420925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,5120,0.03607733382119073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,1024,0.0158933334880405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,768,0.003859555555714501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,768,0.015432889262835184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,4096,0.03357422351837158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,512,0.003567111161020067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,512,0.014888000157144336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,3584,0.03360889024204678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,256,0.0032311110860771606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,256,0.014339556296666464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,3072,0.03180711136923896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,128,0.002974222310715251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,128,0.013917333549923368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,2560,0.03216266632080078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,64,0.0028613333900769553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,2048,0.03020355436537001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,64,0.014031110538376702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,32,0.0028355556229750314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,1536,0.030623111459943984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,32,0.01385600037044949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,1024,0.028954668177498713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,65536,0.018793778287039865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,768,0.029639999071757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,65536,0.17086577415466309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,16384,0.008380444513426887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,16384,0.04693688948949178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,12288,0.007149332927332983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,12288,0.03783555494414435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,10240,0.007237333390447829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,10240,0.032842665910720825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,8192,0.007151111132568783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,8192,0.028599109914567735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,512,0.028928889168633357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,7168,0.007136888802051544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,7168,0.026965333355797663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,256,0.027315555347336665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,6144,0.0069351109365622205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,6144,0.02493688960870107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,128,0.02596355477968852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,5120,0.007296889192528195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,5120,0.02296977738539378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,65536,0.1981226603190104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,4096,0.0069439998931354946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,4096,0.021273778544531927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,16384,0.06392621994018555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,3584,0.007343110938866933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,3584,0.020210666788948905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,10240,0.04981688989533318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,3072,0.0069822221994400024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,3072,0.019115555617544387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,12288,0.053371555275387235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,2560,0.006311999840868845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,2560,0.01844800015290578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,8192,0.04268088936805725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,2048,0.005477333234416113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,2048,0.017313778400421143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,7168,0.040581332312689886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,1536,0.004875555634498596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,1536,0.01644622286160787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,6144,0.03744533326890733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,1024,0.0041884444653987885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,1024,0.015663110547595553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,5120,0.0358222226301829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,768,0.0038071109188927543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,768,0.015255999233987598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,4096,0.033498668008380465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,512,0.0035502223504914176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,512,0.014884443746672736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,3584,0.03385511040687561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,256,0.0031715554909573663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,256,0.014486221803559197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,3072,0.03167288833194309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,128,0.002970666728085942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,128,0.013791110780504016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,2560,0.03209244542651706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,64,0.0028133332315418455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,64,0.013910222384664746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,32,0.002826666666401757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,32,0.013728000223636627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,65536,0.013051555388503604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,65536,0.17226666874355742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,16384,0.007093333535724216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,16384,0.04673155479960971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,12288,0.006976000136799282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,12288,0.037294222248925105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,10240,0.007184888753626082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,10240,0.03313866588804457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,8192,0.007184888753626082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,8192,0.028691556718614366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,7168,0.006963555183675554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,7168,0.02680088910791609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,6144,0.006788444601827198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,6144,0.024887111451890733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,5120,0.007142222589916653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,5120,0.02295822236273024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,4096,0.00673333348499404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,4096,0.02110133402877384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,3584,0.007227555745177799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,2048,0.030200001266267564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,3584,0.020191111498408847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,3072,0.006883555402358373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,2560,0.006147555592987273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,3072,0.019184889064894784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,2560,0.018231110440360177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,2048,0.005453333258628845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,1536,0.004819555415047539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,2048,0.017329777280489605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,1024,0.0041146667467223275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,1536,0.016365332735909354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,1024,0.01555288831392924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,768,0.0037831109431054858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,768,0.015190222197108798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,512,0.003478222423129611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,256,0.003128888913326793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,512,0.014750222365061441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,256,0.014197332991494073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,128,0.002951111023624738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,128,0.013739555246300168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,64,0.002793777734041214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,64,0.01368799971209632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,32,0.0028195555011431375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,32,0.013615111509958902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,65536,0.011811555259757571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,16384,0.007000889215204451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,65536,0.17189866966671416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,1536,0.030825777186287776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,16384,0.04543200135231018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,12288,0.006870222174459034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,1024,0.028484443823496502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,12288,0.0369466675652398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,10240,0.0069955554273393415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,10240,0.033019555939568415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,8192,0.006928000185224745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,8192,0.02863022353914049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,7168,0.0068488890926043195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,7168,0.026909333136346605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,6144,0.00674311113026407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,5120,0.007111111448870764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,6144,0.024987555212444727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,5120,0.02291999922858344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,4096,0.006507555643717448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,4096,0.021181333396169875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,3584,0.006931555353932911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,3584,0.02020355562369029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,3072,0.006712889091836081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,2560,0.006152888966931238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,3072,0.019146665930747986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,2560,0.01830400029818217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,2048,0.005482666608360078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,2048,0.017285333739386666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,1536,0.004821333206362194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,1536,0.016362667083740234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,1024,0.004211555752489301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,1024,0.015563555889659457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,768,0.0037786666717794207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,768,0.029339555237028334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,768,0.015160888433456421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,512,0.0035546666218174826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,512,0.014680888917711047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,256,0.003104000041882197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,64,0.0027555556346972785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,256,0.014294221997261047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,128,0.002944888960984018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,128,0.013811555173661975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,32,0.002760888801680671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,64,0.013744889034165276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,32,0.013696889082590738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,512,0.028792887926101685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,256,0.027541332774692114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,12288,0.3142791059282091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,12288,0.3690799872080485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,16384,0.44002577993604874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,16384,0.40768710772196454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,12288,0.5792284541659886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,10240,0.2627528773413764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,10240,0.4812773068745931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,8192,0.21379644340938994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,16384,0.7600551181369357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,10240,0.31197865804036456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,8192,0.25556710031297475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,7168,0.1944577826393975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,8192,0.38835822211371523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,128,0.02590488890806834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,7168,0.22731733322143555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,6144,0.16512266794840494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,6144,0.294960896174113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,6144,0.20110044214460585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,5120,0.14733600616455078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,7168,0.33910666571723086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,5120,0.1747973362604777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,4096,0.11675111452738444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,3584,0.10478311114841038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,4096,0.1461848947736952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,5120,0.2458284431033664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,3584,0.13513867060343424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,3584,0.17467999458312988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,3072,0.09234133031633164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,4096,0.19951911767323813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,3072,0.15140088399251303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,3072,0.11510222487979466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,2560,0.08019200298521253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,2048,0.0679911110136244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,2560,0.12737954987419978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,1536,0.05589422252443102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,2560,0.10032533274756537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,2048,0.0934986670811971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,1024,0.04417422082689074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,2048,0.10397066672643025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,1536,0.08012266953786214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,1536,0.07490933603710599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,1024,0.06241244739956326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,512,0.04936000042491489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,512,0.03429422113630507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,256,0.026213333010673523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,768,0.0384133325682746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,128,0.01056177748574151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,768,0.056279109583960645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,256,0.01853511068556044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,128,0.02216533323129018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,256,0.040672888358434044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,512,0.032403555181291364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,64,0.008266666697131263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,768,0.04400888747639126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,32,0.007365333537260692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,64,0.025103999508751765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,128,0.036109334892696805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,1024,0.05611733595530192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,32,0.024476443727811176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,12288,0.09919822216033936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,16384,0.12486044565836589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,16384,0.1541395584742228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,12288,0.11679911613464355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,12288,0.14273511038886175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,10240,0.08369866344663833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,16384,0.18803289201524523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,10240,0.10325688785976833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,8192,0.0701084468099806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,10240,0.12015822198655869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,8192,0.09708799918492635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,8192,0.09148799710803562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,7168,0.0633102191819085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,7168,0.0853804416126675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,6144,0.057003557682037354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,7168,0.0800666676627265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,65536,0.4418417877621121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,6144,0.07332888576719496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,6144,0.07484800285763211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,5120,0.06260355313618977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,5120,0.04999822378158569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,4096,0.04240622123082479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,5120,0.06732444630728827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,4096,0.051643557018703885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,4096,0.05911733044518364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,3584,0.03958311014705234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,3584,0.057111998399098717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,3584,0.04562577936384413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,3072,0.03928000066015456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,3072,0.04023111197683547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,2560,0.03375377919938829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,3072,0.05594578054216173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,2560,0.051183998584747314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,2048,0.03018310997221205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,2048,0.028471999698215063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,2560,0.034239110019471906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,1536,0.02599999970859951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,1536,0.04192622171507942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,1024,0.021792888641357422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,768,0.012479999827014076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,2048,0.047224889198939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,1024,0.038324445486068726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,768,0.020818667279349435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,512,0.009359999663299983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,1024,0.01575911045074463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,768,0.03710133168432448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,512,0.019327110714382596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,1536,0.0226248883538776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,256,0.00655644428398874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,256,0.015871110889646742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,512,0.03459111187193129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,128,0.004307555655638377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,256,0.029283556673261855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,128,0.01519377695189582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,128,0.027116444375779893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,64,0.003936000168323517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,32,0.0041288890772395665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,64,0.015871110889646742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,32,0.015885333220163982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,65536,0.49176888995700413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,16384,0.10911733574337429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,16384,0.12479378117455377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,16384,0.15670400195651585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,12288,0.08616799778408474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,12288,0.10967466566297744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,12288,0.10212977727254231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,10240,0.07288177808125813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,65536,0.7723644574483236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,10240,0.09439555803934734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,10240,0.0968559980392456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,8192,0.061358220047420926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,8192,0.07551911142137316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,8192,0.08146044280793932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,7168,0.055326223373413086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,65536,0.3794479899936252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,7168,0.06912355290518866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,6144,0.04962755574120415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,7168,0.07110222180684407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,5120,0.044222222434149853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,6144,0.06626311275694105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,5120,0.049288001325395375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,65536,0.44428711467319065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,4096,0.038367112477620445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,6144,0.05829155445098877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,5120,0.06091644366582235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,4096,0.05352266629536947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,4096,0.039261334472232394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,3584,0.03808533483081394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,3584,0.034847110509872437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,3584,0.05472977624999153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,3072,0.03233511249224345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,2560,0.03039466672473484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,2048,0.02607911162906223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,3072,0.048735111951828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,2560,0.04751911097102695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,2048,0.04278666774431864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,1536,0.023379556006855432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,1024,0.012261333564917246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,3072,0.030949334303538006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,1536,0.03898577888806661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,1024,0.020994666549894545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,2048,0.02211911148495144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,1536,0.01757244434621599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,1024,0.036919110351138644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,2560,0.026480888326962788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,768,0.010242666635248397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,768,0.019997333486874897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,512,0.008058666355080074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,768,0.03591555688116286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,512,0.01841600073708428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,256,0.005510222166776657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,256,0.015678221980730694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,512,0.034190220965279475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,128,0.003991111285156674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,256,0.028837333122889202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,128,0.015150222513410779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,64,0.003583999971548716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,128,0.02757688860098521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,32,0.0038577777643998465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,64,0.015625778171751235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,32,0.015359110302395292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,16384,0.10035110844506158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,16384,0.11398488945431179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,12288,0.09095911184946696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,12288,0.08034577634599474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,16384,0.1383777724372016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,12288,0.09427377912733291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,10240,0.0681413345866733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,10240,0.08325066831376818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,8192,0.06158044603135851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,10240,0.07648533582687378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,8192,0.0572835538122389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,65536,0.34798489676581484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,8192,0.07126222054163615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,7168,0.05167288912667168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,65536,0.572207980685764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,7168,0.054676443338394165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,7168,0.06705244382222493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,6144,0.04704533351792229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,6144,0.06253600120544434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,6144,0.048397332429885864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,5120,0.04213066564665901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,4096,0.03672000103526645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,5120,0.058077335357666016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,4096,0.0520008901755015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,3584,0.03363555669784546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,4096,0.038239111502965294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,5120,0.040884445110956825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,3584,0.030006223254733618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,3072,0.030841777722040813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,3072,0.04643911123275757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,2560,0.028239998552534316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,2560,0.044687112172444664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,2048,0.024783111280865137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,3584,0.050632887416415744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,3072,0.026808000273174707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,1536,0.014944000376595391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,2048,0.04085155659251743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,1024,0.010913777682516309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,1536,0.022487999664412603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,1536,0.0376506679587894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,2048,0.019141332970725167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,2560,0.025851555996470984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,1024,0.020374221934212577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,768,0.009233777721722921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,1024,0.036892443895339966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,65536,0.39769334263271755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,768,0.019392000304328073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,768,0.03513866662979126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,512,0.007535999847782983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,512,0.017095999585257638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,256,0.004780444420046276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,512,0.03264355659484863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,256,0.015769778026474845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,256,0.028254220883051556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,128,0.0039822223285834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,128,0.015134221977657743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,64,0.003597333199448056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,128,0.026605332891146343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,32,0.0037679999238914917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,64,0.015525332755512662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,32,0.015101333459218344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,65536,0.4956800142923991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,16384,0.07928622431225248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,12288,0.07384000221888225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,16384,0.09917155239317153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,16384,0.1047546664873759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,12288,0.06996622350480822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,12288,0.08324888679716322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,10240,0.054888000090916954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,10240,0.07326844665739271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,8192,0.05051911208364698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,10240,0.06201066573460897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,8192,0.04673244555791219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,7168,0.04284977912902832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,8192,0.06323111057281494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,7168,0.04453066653675503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,7168,0.05986577934688992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,6144,0.03909066650602553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,5120,0.03509777784347534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,6144,0.05593244234720866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,6144,0.03936622209019131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,5120,0.03347555465168423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,5120,0.05214666657977634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,4096,0.031144890520307753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,4096,0.027446221974160936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,3584,0.028914666838116113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,4096,0.04654399885071648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,3584,0.04505599869622124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,3584,0.024938666158252295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,65536,0.3228151003519694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,3072,0.026410667432679072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,65536,0.2585679954952664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,3072,0.042196445994906955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,3072,0.021810667382346258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,2560,0.02442844377623664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,2560,0.018724444839689467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,2560,0.04064888755480448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,2048,0.02239111065864563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,1536,0.013660444153679742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,1536,0.02089244458410475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,2048,0.03903111153178745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,1024,0.009143110778596666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,1536,0.0361057784822252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,1024,0.01921688848071628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,768,0.007707555260923173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,768,0.01759377784199185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,1024,0.03555466731389364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,512,0.006315555423498154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,2048,0.017243555850452848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,512,0.01625866691271464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,768,0.03453866640726725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,256,0.004206222378545337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,512,0.031162665949927434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,256,0.014992000328169929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,128,0.0038115556041399636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,256,0.028119110398822363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,128,0.014495111174053617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,32,0.014529777897728814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,128,0.026464000344276428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,64,0.0034302222645945023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,64,0.01460533340771993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,32,0.003539555602603488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,16384,0.07348889112472534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,16384,0.09419911437564427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,12288,0.06517599688635932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,16384,0.09284800291061401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,12288,0.06008711126115587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,10240,0.05070933368470934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,12288,0.07875821987787883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,10240,0.06996711095174153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,10240,0.05496177739567227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,65536,0.4089964495764838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,8192,0.043976889716254346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,8192,0.0603733327653673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,7168,0.04342222213745117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,8192,0.04437866806983948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,7168,0.03944799966282315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,7168,0.05724355247285631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,6144,0.03684799869855245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,6144,0.05367288986841837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,6144,0.03521155648761325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,4096,0.029190222422281902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,5120,0.029932445949978296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,5120,0.033195555210113525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,65536,0.2378337648179796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,4096,0.04490310947100321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,5120,0.05044355657365587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,3584,0.027471999327341717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,4096,0.024656888511445787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,3072,0.024767110745112102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,3584,0.044144001271989614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,3072,0.040977779361936785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,3584,0.022113778524928625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,2560,0.023363555471102398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,2048,0.013632000320487551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,2560,0.040623999304241605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,3072,0.019632889164818656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,2048,0.021744888689782884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,2560,0.01677600045998891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,1536,0.011016888750924004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,2048,0.038402666648228966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,1536,0.02051111062367757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,1024,0.008565333154466417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,1024,0.0173342227935791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,768,0.00721777809990777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,1536,0.036362667878468834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,768,0.016861332787407767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,512,0.0057946667075157166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,1024,0.03613155417972141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,512,0.01568533314598931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,768,0.034653332498338484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,256,0.003955555458863576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,512,0.029696888393825952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,128,0.003589333345492681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,256,0.014856000741322836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,256,0.02814488940768772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,128,0.014316444595654806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,65536,0.29676710234747994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,64,0.003565333369705412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,128,0.026921777261628047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,64,0.014530667000346713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,32,0.003620444486538569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,32,0.014372444815105863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,16384,0.0683786670366923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,65536,0.3596506648593479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,16384,0.09503999683592056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,16384,0.08511022064420913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,12288,0.060290667745802135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,12288,0.05693866809209188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,12288,0.07987733019722833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,10240,0.04769688844680786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,10240,0.07092355357276069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,8192,0.041136887338426374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,10240,0.048102223210864596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,8192,0.038860443565580584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,8192,0.061202665170033775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,7168,0.03481333454449972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,6144,0.03495733274353875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,7168,0.03802844550874498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,7168,0.05800444549984402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,65536,0.21867733531528047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,6144,0.030863109562132094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,5120,0.031347556246651545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,5120,0.05012177758746677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,65536,0.2917439937591553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,4096,0.027431999643643696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,6144,0.05413422319624159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,4096,0.04475555486149258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,3584,0.02568799919552273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,3584,0.04313155677583483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,3072,0.023964444796244305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,5120,0.026361778378486633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,2560,0.014724445011880664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,3072,0.04067111015319824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,4096,0.021770666042963665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,2560,0.022740445203251306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,3072,0.017156443662113614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,3584,0.019515555765893724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,2048,0.012460444536474017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,2560,0.04064977831310696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,2048,0.021272889441914026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,1536,0.009876444107956356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,2048,0.03766311208407084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,1536,0.019689778486887615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,1024,0.007977777885066139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,1536,0.03546222382121616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,1024,0.017419555120997958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,768,0.006780444333950679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,1024,0.03454222281773885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,768,0.017251556118329365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,512,0.004811555561092165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,768,0.033202668031056724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,512,0.015641777051819693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,256,0.003882666842805015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,512,0.030760000149408977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,256,0.014767999450365702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,128,0.003531555748648114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,256,0.02773422168360816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,128,0.014159111513031853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,128,0.026447110705905493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,64,0.0034666665726237824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,64,0.014684443672498068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,32,0.0035048888789282907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,32,0.014208889669842191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,65536,0.2960382302602132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,12288,0.05225777626037598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,16384,0.06298133399751452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,16384,0.08260889185799493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,12288,0.05621688895755344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,10240,0.04483377933502197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,12288,0.07016355461544461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,16384,0.07111466593212552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,10240,0.06275288926230536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,10240,0.04122044311629401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,8192,0.03890044490496317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,7168,0.03596355517705282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,8192,0.033258666594823204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,8192,0.05437689026196798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,6144,0.03276355399025811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,7168,0.03006311257680257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,7168,0.051875554853015475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,6144,0.049031111929151744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,5120,0.029346668057971533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,5120,0.04569333460595873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,4096,0.0255724440018336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,65536,0.2039768828286065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,5120,0.022804444034894306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,6144,0.027063111464182537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,4096,0.04096444447835287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,3584,0.024355555574099224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,4096,0.018819555640220642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,3072,0.014771555860837301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,3584,0.040341334210501775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,3072,0.023316444622145757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,3584,0.016889777448442247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,2560,0.012585777375433179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,3072,0.03843377696143256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,2560,0.021827555365032617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,2048,0.01074488874938753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,2560,0.03825599948565165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,1536,0.008916444248623317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,2048,0.02076088885466258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,1536,0.019320888651741874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,2048,0.036552889479531184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,1024,0.007098666495747036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,1536,0.034845332304636635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,1024,0.017699556218253244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,768,0.006143110907740063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,1024,0.0348639984925588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,768,0.01607733302646213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,512,0.004261333495378494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,768,0.03155555658870273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,512,0.015649777319696214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,256,0.0037831109431054858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,512,0.02964444292916192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,256,0.01477688882086012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,128,0.0034657776769664553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,256,0.027494221925735474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,128,0.014110222458839417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,65536,0.2671644422743055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,64,0.0032293332947625052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,32,0.003343111111058129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,128,0.02672622270054287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,64,0.014392000105645923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,32,0.014134221606784396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,16384,0.0610062215063307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,16384,0.07667999797397189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,16384,0.0558640029695299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,12288,0.04982399940490723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,12288,0.0652737749947442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,12288,0.044630222850375705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,10240,0.04293955696953667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,10240,0.05915288792716133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,65536,0.1973795493443807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,8192,0.03235822253757053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,65536,0.24417599042256674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,10240,0.03856533434655931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,7168,0.03405244482888116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,8192,0.03722666700681051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,8192,0.05132977830039131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,7168,0.04926133486959669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,6144,0.031074666314654883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,6144,0.04678666591644287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,5120,0.027991112735536363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,7168,0.029578665892283123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,5120,0.043530666165881686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,65536,0.24577334192064074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,4096,0.025447999437650044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,3584,0.01367911116944419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,5120,0.018921777606010437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,3584,0.024134222004148696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,4096,0.03934489025010003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,3072,0.012242666549152799
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,6144,0.026691555976867676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,3584,0.038959999879201256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,3072,0.022693332698610093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,2560,0.010786666638321347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,2560,0.021497777766651575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,4096,0.015602666470739575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,3072,0.03717599974738227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,2048,0.009286222358544668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,2048,0.020635555187861126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,2560,0.03765155540572272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,1536,0.00797333319981893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,2048,0.03599377804332309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,1536,0.018920888503392536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,1536,0.03389244609408908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,1024,0.007052444749408298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,1024,0.01661688917213016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,768,0.004923555586073133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,1024,0.031672000885009766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,768,0.016043555405404832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,512,0.004296888907750447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,768,0.030373334884643555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,512,0.01571822166442871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,256,0.0035742223262786865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,512,0.02878755662176344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,256,0.014785778191354541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,128,0.0032915555768542816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,256,0.02730400032467312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,128,0.014241778188281588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,64,0.003128000017669466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,32,0.014030221435758801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,128,0.026018665896521673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,64,0.014200000299347771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,32,0.00349866681628757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,16384,0.057612445619371205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,16384,0.0734871096081204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,16384,0.0499253339237637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,65536,0.19931556118859184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,12288,0.04049955474005805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,12288,0.04808355702294243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,10240,0.05746666590372721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,12288,0.06436266501744588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,8192,0.03443911009364658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,10240,0.04393511017163595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,8192,0.05049688948525322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,10240,0.03465777635574341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,7168,0.03275377882851495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,7168,0.04838933216200935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,6144,0.02973777718014187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,8192,0.029342220889197454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,6144,0.04579288760821024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,7168,0.02388800018363529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,6144,0.021055999729368422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,5120,0.02700177828470866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,5120,0.042783111333847046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,4096,0.013909333282046847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,4096,0.02426311042573717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,5120,0.016795555750528973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,3584,0.012558221817016602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,4096,0.03956355651219686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,3584,0.023168888356950548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,3072,0.011253332926167382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,65536,0.18543644746144614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,3584,0.03837066557672288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,65536,0.23334044880337187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,3072,0.021856889128684998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,2560,0.010024888647927178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,3072,0.036779556009504534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,2560,0.021197333931922913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,2048,0.008681777450773451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,2048,0.01982933282852173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,2560,0.03680355681313409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,1536,0.007287110719415877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,2048,0.03539289037386576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,1536,0.01735822194152408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,1024,0.006256888724035687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,1024,0.016607999801635742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,1536,0.033583998680114746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,768,0.005180444568395615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,1024,0.030922667847739324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,768,0.015947555502255756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,512,0.003930666794379552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,768,0.0303457776705424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,512,0.015405333704418607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,256,0.003485333174467087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,512,0.029000888268152874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,256,0.014747555057207743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,128,0.0033155555526415506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,256,0.02730133301681942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,128,0.014335110783576965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,128,0.026184888349639043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,64,0.003120888852410846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,65536,0.18181778324974907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,32,0.0032764443506797156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,64,0.014124444789356656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,32,0.013984888792037964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,16384,0.056295997566647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,12288,0.046559111939536206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,16384,0.07359644439485338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,12288,0.035070220629374184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,16384,0.0469742218653361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,10240,0.03971911138958401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,12288,0.0628604425324334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,10240,0.05688889159096611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,8192,0.03456977672047085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,7168,0.03207733233769735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,8192,0.04860622353023953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,10240,0.030896888838873968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,8192,0.026120889517996047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,7168,0.04753777715894911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,7168,0.02201333310869005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,6144,0.029270221789677937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,5120,0.015108444624476962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,6144,0.045216000742382474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,5120,0.02652799917591943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,6144,0.01794399983353085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,4096,0.012431110772821637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,4096,0.024114666713608637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,5120,0.04300622145334879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,3584,0.011393778026103973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,4096,0.03858044412400987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,3584,0.023390221926901076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,3072,0.010424888796276517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,3584,0.038303110334608294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,3072,0.022052443689770166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,2560,0.009287111461162567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,2560,0.020945777495702107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,3072,0.03648799988958571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,2048,0.008079110748238033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,65536,0.16835467020670572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,2048,0.019715555840068393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,2560,0.036863999234305486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,1536,0.006970666348934174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,2048,0.03479911221398248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,65536,0.22778399785359701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,1536,0.017143999536832173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,1536,0.0331928895579444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,1024,0.005673777725961473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,1024,0.01628444426589542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,768,0.004563555535342959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,1024,0.030919111437267725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,768,0.01578044394652049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,768,0.02980088856485155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,512,0.003999999827808804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,512,0.015453333655993143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,256,0.00350222239891688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,512,0.028631998433007136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,256,0.014737778239780001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,256,0.02737244466940562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,128,0.0033075554917256036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,128,0.014352000421947904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,64,0.0031306667046414483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,32,0.0032195556494924757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,128,0.026384888423813715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,64,0.014027555783589682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,32,0.013935999737845527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,65536,0.15221422248416477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,16384,0.053782221343782216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,16384,0.07170933485031128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,12288,0.04397955536842346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,12288,0.06340799729029338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,16384,0.03784444597032335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,10240,0.03702577617433336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,12288,0.030533333619435627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,10240,0.055958224667443164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,8192,0.03259822063975864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,10240,0.026904000176323786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,8192,0.04951111144489712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,8192,0.022517333428064983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,7168,0.030258665482203167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,6144,0.015632000234391954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,7168,0.04673333300484551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,6144,0.027702222267786663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,7168,0.017695110705163743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,5120,0.013307555682129331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,65536,0.16949778132968477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,6144,0.04546844296985202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,5120,0.025536000728607178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,4096,0.011090666883521609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,4096,0.02363555630048116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,5120,0.04161955581771003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,3584,0.010337777435779572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,3584,0.022388445006476507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,4096,0.038310223155551486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,3072,0.010472888747851053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,3584,0.03808533483081394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,3072,0.021324444148275588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,2560,0.008657777474986183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,2560,0.020375111036830477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,65536,0.21765955289204916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,3072,0.03618489040268792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,2560,0.03680177860789829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,2048,0.0074471111098925276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,2048,0.01901155544651879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,1536,0.006737777756320105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,2048,0.03457866774664985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,1536,0.016918222109476726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,1024,0.005631999837027655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,1536,0.03198488884501987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,1024,0.01612355477280087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,768,0.004661333229806689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,1024,0.03035022152794732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,768,0.015607110328144498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,512,0.003951999876234266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,768,0.030062221818500098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,512,0.015297777122921415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,512,0.028805332051383123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,256,0.0034799998005231223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,256,0.014456000592973499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,128,0.0032515554792351196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,256,0.0273635552989112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,64,0.00304888892504904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,128,0.014223110344674854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,65536,0.12931822405921087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,128,0.026193777720133465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,64,0.01387644476360745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,32,0.0031155554784668815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,32,0.013936888840463428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,12288,0.04213866591453552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,16384,0.05168888966242472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,12288,0.025478222303920325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,12288,0.06083644760979546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,16384,0.068686220380995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,10240,0.03626311156484816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,16384,0.032032890452278986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,10240,0.05402222275733948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,8192,0.03161244591077169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,7168,0.016160888804329764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,8192,0.04637155599064297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,7168,0.029873778422673542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,10240,0.02254666719171736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,7168,0.04489955637190077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,6144,0.014192000031471252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,8192,0.01884533299340142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,6144,0.027309333284695942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,5120,0.012615111139085559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,6144,0.043418665726979576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,5120,0.024449777272012498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,4096,0.010825778047243753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,65536,0.16633689403533936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,4096,0.023413333627912734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,3584,0.009648000200589498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,5120,0.04164266586303711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,4096,0.03739022215207418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,3584,0.022105778257052105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,3584,0.037426667080985174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,3072,0.008984000318580205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,3072,0.019613333874278598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,2560,0.008177777959240807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,65536,0.20564622349209258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,3072,0.03561688794030084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,2560,0.01993777851263682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,2560,0.035475555393430926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,2048,0.007252444823582967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,1536,0.01683022247420417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,2048,0.01759377784199185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,1536,0.0064382221963670515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,2048,0.032280001375410296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,1024,0.005041777673694822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,1536,0.031009776724709406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,1024,0.01604711181587643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,768,0.004212444441186057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,1024,0.03093866507212321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,768,0.015712888704405892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,512,0.0037697777152061462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,768,0.02997066577275594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,256,0.01462222304609087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,512,0.015136889285511441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,256,0.003411555455790626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,512,0.028679112593332928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,65536,0.1063235534562005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,128,0.003198222153716617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,256,0.02757511039574941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,128,0.014168888330459595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,64,0.003028444531891081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,128,0.025920889443821375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,64,0.013970666461520724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,32,0.003149333306484752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,32,0.013897778259383308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,16384,0.04917422268125746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,16384,0.06385688649283515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,12288,0.03921066721280416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,10240,0.018610666195551556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,16384,0.026111111044883728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,12288,0.05711999866697523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,10240,0.03418666786617703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,8192,0.015285332997639975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,12288,0.021185777253574792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,10240,0.05025510986646017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,7168,0.013863999810483722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,8192,0.029828445778952703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,7168,0.028111110130945843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,8192,0.04402755697568258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,6144,0.012462221913867526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,65536,0.15481332937876383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,7168,0.04296888907750448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,5120,0.011179555621412067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,6144,0.026378666361172993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,5120,0.024414221445719402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,6144,0.04188977678616842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,4096,0.009713778065310584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,5120,0.03940533267127143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,4096,0.022304889228608873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,3584,0.009015999734401703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,3584,0.02138311167558034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,3584,0.03684622380468581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,4096,0.03675999906327989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,3072,0.008347555167145198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,3072,0.019496889577971566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,2560,0.007651555869314406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,3072,0.03440622157520718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,2560,0.01902933253182305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,2048,0.006809777683681912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,65536,0.1838364468680488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,2560,0.03369777732425266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,2048,0.017456889152526855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,1536,0.005571555760171678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,2048,0.03190755512979295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,1536,0.016851555969980028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,1024,0.004679999831649992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,1536,0.030400888787375555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,65536,0.08363022406895955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,1024,0.015976889265908133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,768,0.00434755575325754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,1024,0.030343112018373277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,768,0.015459555718633862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,512,0.0038195554580953387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,256,0.014467555615637036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,768,0.03007377849684821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,512,0.015104000767072042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,256,0.0034097778714365433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,512,0.02858933475282457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,128,0.003120888852410846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,256,0.02719377809100681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,64,0.0029964444951878656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,128,0.01424888935354021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,32,0.003072000005178981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,64,0.014053333136770459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,128,0.026199999782774184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,32,0.013738666971524557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,16384,0.019377777973810833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,16384,0.04626577761438158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,12288,0.01567911108334859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,12288,0.037490665912628174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,16384,0.06308355596330431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,10240,0.014783110883500842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,10240,0.03323377834426032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,12288,0.056442664729224316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,10240,0.05006844467586941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,8192,0.013182222015327878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,8192,0.02944799926545885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,7168,0.01221333361334271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,8192,0.044794665442572705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,65536,0.1500613292058309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,7168,0.027875555886162653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,6144,0.011326221956147088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,7168,0.042902221282323204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,6144,0.02602311140961117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,5120,0.010587555666764578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,6144,0.04156000084347195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,5120,0.023080888721677992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,4096,0.009579555855857002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,65536,0.18334578143225777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,5120,0.038768000072903104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,4096,0.02145511077509986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,65536,0.06381599770651923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,3072,0.008017777568764156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,4096,0.034372445609834455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,3584,0.00867733359336853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,3584,0.02028888960679372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,3584,0.03477599885728624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,3072,0.019275556008021038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,2048,0.005869333528810077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,2560,0.0069057775868309875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,3072,0.03306666678852505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,2560,0.018214222457673814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,2560,0.033384889364242554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,1536,0.00499555551343494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,2048,0.017551110850440133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,2048,0.032171554035610624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,1536,0.016719111137919955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,1024,0.004276444514592488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,1536,0.02983199887805515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,1024,0.016009777784347534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,768,0.003957333250178231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,768,0.015583111180199517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,1024,0.030751112434599135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,512,0.003592888928121991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,768,0.02979822291268243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,512,0.014955555399258932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,256,0.003272888975010978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,512,0.028631998433007136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,256,0.014506666196717156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,128,0.0030826667530669104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,256,0.02699022160636054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,64,0.0028977776981062363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,128,0.013944889108339945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,32,0.002966222249799304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,128,0.025826666090223525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,64,0.01384888920519087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,32,0.013798221945762634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,16384,0.01772799922360314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,65536,0.14994133843315974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,16384,0.04485866758558485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,12288,0.014919110470347934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,16384,0.05916088819503784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,12288,0.0380675560898251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,10240,0.03279111120435927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,12288,0.052601777844958834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,10240,0.01351199961370892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,8192,0.012120000190205045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,65536,0.17074844572279188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,10240,0.04857155680656433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,8192,0.029326223664813574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,8192,0.04261066516240438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,7168,0.011300444602966309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,7168,0.027691556347741023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,6144,0.010640000303586325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,7168,0.04150311152140299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,6144,0.025647999511824712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,65536,0.05123555660247803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,6144,0.03952888978852166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,5120,0.009925333162148794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,5120,0.02271911170747545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,4096,0.008544888761308458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,5120,0.03789600067668491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,4096,0.021345777644051447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,4096,0.033366223176320396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,3584,0.008288000192907121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,3584,0.02052088909678989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,3584,0.034528887934154935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,3072,0.007711110843552484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,3072,0.019105777144432068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,2048,0.017439999514155917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,2560,0.006783999916579988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,1536,0.004964444372389051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,3072,0.03202844328350491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,2560,0.018254222141371835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,2048,0.00590133335855272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,2560,0.03314133485158285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,2048,0.031383110417260066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,1536,0.016805332567956712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,1024,0.004322666674852371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,1024,0.01608622239695655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,1536,0.03001333276430766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,768,0.003976888954639435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,768,0.015555555621782938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,1024,0.030448890394634668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,512,0.028361777464548748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,768,0.02937244375546773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,512,0.0036391110883818734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,512,0.015031110909250049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,256,0.0032559999575217566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,128,0.013928888572586907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,256,0.014416888356208801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,128,0.003039111072818438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,256,0.027227555712064106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,64,0.0028826666788922418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,32,0.013817778064144982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,128,0.02645066711637709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,32,0.0029191111938820947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,64,0.013820444544156393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,16384,0.0140328887436125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,16384,0.04421066575580173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,12288,0.012124444047609964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,12288,0.03709422217475043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,16384,0.05849599838256836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,10240,0.011262222296661802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,12288,0.0521342224544949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,10240,0.032123555739720665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,8192,0.010266666611035665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,8192,0.028716444969177246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,7168,0.009775111244784461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,10240,0.0473288893699646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,65536,0.14633333683013916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,8192,0.041815112034479775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,65536,0.16318133142259386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,7168,0.026704000102149114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,65536,0.038019554482565984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,7168,0.04093066520161099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,6144,0.008965333302815756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,6144,0.024767110745112102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,5120,0.008065777520338694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,6144,0.038000000847710505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,5120,0.022077333596017625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,5120,0.035939554373423256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,4096,0.007620444728268518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,4096,0.02053955528471205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,3584,0.0075502221783002215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,4096,0.03341866532961527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,3584,0.019923556182119582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,3072,0.007194666398896112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,3584,0.03388177686267429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,3072,0.018970666660202872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,2560,0.006367110957702001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,3072,0.03206844462288751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,2560,0.01811644434928894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,2560,0.03257511059443156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,2048,0.005675555517276128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,2048,0.017289777596791584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,1536,0.004977777600288391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,2048,0.03143822153409322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,768,0.003917333152559069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,1536,0.016553777787420485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,1024,0.004297777596447202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,1536,0.030335999197430078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,1024,0.015989333391189575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,1024,0.029990222718980577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,768,0.01536355581548479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,512,0.0035537779331207275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,768,0.029762665430704754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,512,0.014911999305089315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,256,0.0032275555034478507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,512,0.028215110301971436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,256,0.014383110735151501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,256,0.026914666096369427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,128,0.002990222225586573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,128,0.013841778039932251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,64,0.00286666676402092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,128,0.026332444614834253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,64,0.01385600037044949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,32,0.002858666703104973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,32,0.013822221921549903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,16384,0.010418666733635796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,65536,0.1439946624967787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,16384,0.04383644461631775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,12288,0.009305777649084726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,65536,0.16346932782067194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,65536,0.024305777417288885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,16384,0.058196445306142174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,12288,0.03622577918900384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,10240,0.008316444853941599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,12288,0.05133688780996534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,10240,0.03126044405831231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,8192,0.0075635554061995606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,10240,0.046258668104807533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,6144,0.006953777538405524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,8192,0.02772177755832672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,7168,0.007243555453088548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,8192,0.041229334142473005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,7168,0.02570488883389367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,6144,0.023726221587922838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,7168,0.039127998881869845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,5120,0.007418666448858049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,5120,0.022068444225523207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,6144,0.0373982224199507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,5120,0.03563111027081808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,4096,0.006955555743641323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,4096,0.02033866610791948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,3584,0.007339555356237624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,4096,0.0328328874376085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,3584,0.01964000033007728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,3072,0.007096889118353526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,3584,0.033215999603271484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,3072,0.018696889281272888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,2560,0.006169777777459886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,3072,0.03188622328970168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,2560,0.017808000246683758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,2048,0.005538666827811136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,2560,0.032051556640201144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,2048,0.017113778326246474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,1536,0.004839999808205498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,2048,0.031179553932613794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,1536,0.016238222519556682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,1024,0.004210666649871402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,1536,0.030108445220523413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,1024,0.015804444750150044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,768,0.003997333347797394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,1024,0.02980977627966139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,768,0.015214223000738354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,512,0.003552888830502828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,768,0.029024001624849107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,512,0.014674666855070325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,256,0.003170666595300039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,256,0.014236445228258768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,512,0.028388443920347426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,128,0.0030293334275484085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,256,0.027181333965725366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,64,0.0028071111689011254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,128,0.01387555566098955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,128,0.025860443711280823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,64,0.013904000322024027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,32,0.002854222224818336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,32,0.013713777893119387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,65536,0.143267552057902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,65536,0.015942222542232938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,12288,0.007180444068378872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,16384,0.008016000191370646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,16384,0.0426879988776313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,12288,0.05129244592454699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,65536,0.16421511438157824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,16384,0.05787999762429131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,10240,0.0073315559162033936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,12288,0.03465777635574341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,10240,0.030799110730489094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,8192,0.007137777904669444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,10240,0.045238223340776235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,8192,0.027052443888452318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,7168,0.007058666812049017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,8192,0.03973244296179877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,7168,0.02536800007025401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,6144,0.0069146665434042616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,5120,0.007239999870459239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,7168,0.038422223594453596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,6144,0.02383288906680213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,5120,0.022091555926534865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,4096,0.006836444553401735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,6144,0.037251555257373385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,5120,0.03585422370168898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,4096,0.020303999384244282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,3584,0.007286222444640265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,4096,0.03270577722125583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,3584,0.019484443797005545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,3072,0.03238933285077413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,2560,0.017768000562985737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,2048,0.005513777749405966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,3072,0.006940444310506185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,3584,0.03299644589424133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,3072,0.018631999691327412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,2560,0.006142222219043308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,2560,0.031847112708621554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,2048,0.017052445146772597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,1536,0.029394666353861492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,1536,0.004839111119508743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,2048,0.030812445614073012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,1536,0.016213332613309223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,1024,0.004208000169859992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,1024,0.015496888094478183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,768,0.0037991110649373797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,768,0.01521511044767168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,1024,0.02980355421702067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,512,0.0034719999465677473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,768,0.029309332370758057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,512,0.014661333627170987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,256,0.0031475555151700974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,512,0.02812266680929396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,256,0.014347554908858405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,128,0.0030008889734745026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,256,0.026657778355810378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,128,0.013762666947311826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,64,0.0028248888750871024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,128,0.02565066681967841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,32,0.0028151110228565005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,64,0.013800889253616333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,32,0.013897778259383308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,65536,0.013184000220563678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,16384,0.0070426662762959795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,65536,0.14315022362603083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,12288,0.006994666324721442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,16384,0.04126133190260993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,10240,0.0071937781241205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,12288,0.034627556800842285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,8192,0.007045333584149678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,10240,0.030538668235143025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,7168,0.006968888971540663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,8192,0.027112000518374976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,6144,0.00683111117945777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,7168,0.025310221645567153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,5120,0.007159999675220913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,6144,0.02375999920898014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,4096,0.006768888897365994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,3072,0.0068862222962909276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,5120,0.02204533252451155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,3584,0.00721955547730128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,4096,0.02023555503951179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,3584,0.019369777705934312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,2560,0.006161777923504512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,3072,0.018607111440764535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,2048,0.005495999836259418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,2560,0.017664000391960144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,2048,0.016945777667893302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,1536,0.0047884442740016514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,1536,0.016160888804329764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,1024,0.004152889053026835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,1024,0.01552888916598426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,768,0.003793777690993415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,512,0.0034737777378824023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,768,0.015079110860824585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,128,0.013946667313575745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,256,0.0031306667046414483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,512,0.014693333042992486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,128,0.0029297777348094513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,256,0.014183999763594734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,64,0.00282577777074443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,32,0.002871111035346985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,64,0.013648000028398303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,65536,0.011492444409264458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,32,0.013591999808947245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,16384,0.0070257774657673305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,12288,0.006959111326270633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,65536,0.1429306666056315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,16384,0.04121511181195577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,10240,0.006952889263629913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,12288,0.03457599878311157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,8192,0.0069520001610120135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,10240,0.030361778206295435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,7168,0.006809777683681912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,8192,0.027130666706297133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,6144,0.006710222197903528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,7168,0.02536711096763611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,5120,0.007073777417341868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,6144,0.023590222001075745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,4096,0.006708444406588872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,5120,0.021864889396561518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,3584,0.0070568886068132185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,4096,0.020420443680551317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,3072,0.006688000013430913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,3584,0.01941333380010393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,2560,0.006126222097211414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,3072,0.01865511139233907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,2048,0.005468444277842839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,2560,0.0177848885456721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,1536,0.00479288895924886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,2048,0.01703288820054796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,1024,0.004104888687531154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,1536,0.016135111451148987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,768,0.0037724446091387006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,1024,0.015428443749745687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,512,0.0034622223012977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,768,0.015026667051845126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,256,0.0031795555518733132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,512,0.014712888333532544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,128,0.0029484445436133277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,256,0.014072888427310519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,64,0.0027662221756246355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,128,0.013799111048380533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,32,0.0027742222365405825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,64,0.013737777868906656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,32,0.013833777772055732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,12288,0.31344445546468097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,10240,0.2639111147986518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,12288,0.36339643266465926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,16384,0.4104311201307509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,16384,0.4372604423099094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,8192,0.2136320008171929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,12288,0.5926399760776097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,10240,0.30981243981255424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,10240,0.4924142095777724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,7168,0.1894915501276652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,8192,0.2541626559363471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,8192,0.39791022406684023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,7168,0.22515998946295845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,6144,0.18439111444685194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,7168,0.34589245584275985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,6144,0.29784711201985675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,6144,0.20094222492641875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,5120,0.15749333964453802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,16384,0.7699164284600152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,5120,0.17049154970380995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,4096,0.12941955195532903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,4096,0.14695021841261122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,3584,0.10392533408270942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,4096,0.20260266462961832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,5120,0.25095200538635254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,3072,0.09221600161658393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,3584,0.12380977471669515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,3584,0.17926222748226592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,3072,0.1547351148393419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,3072,0.11258399486541748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,2048,0.07506310939788818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,2560,0.08024266693327162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,2048,0.08560266759660508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,2560,0.1063377790980869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,1536,0.05535377727614509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,2048,0.10485866996977065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,768,0.03759377863672044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,768,0.05492800143029955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,1024,0.04758755697144402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,1536,0.07522755861282349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,1536,0.08272977670033772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,2560,0.13056000073750815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,1024,0.061660442087385386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,768,0.04589510957400004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,512,0.03207555413246155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,256,0.025449777642885845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,128,0.011228444675604502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,128,0.02214755614598592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,256,0.040595554643207125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,64,0.008662222160233391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,1024,0.05663822094599406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,128,0.03547733359866672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,256,0.019305777218606737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,512,0.03277333246337043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,64,0.024570667081409033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,32,0.007287999822033777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,512,0.04892444279458788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,32,0.0247715562582016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,12288,0.09585333532757229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,16384,0.12398310502370198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,16384,0.151110225253635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,12288,0.11867911285824245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,12288,0.1444231139289008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,16384,0.1885155571831597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,10240,0.12054844697316487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,10240,0.08254488971498278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,8192,0.06995377937952678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,10240,0.09988444381290013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,7168,0.06262399752934773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,8192,0.0846595565478007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,7168,0.07893155680762397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,8192,0.0979528890715705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,6144,0.05605600277582804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,7168,0.08583199977874756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,6144,0.07286577754550509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,6144,0.07502310805850558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,5120,0.0491884450117747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,5120,0.06571733289294772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,4096,0.04257066713439094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,5120,0.0634044475025601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,4096,0.05942933426962959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,4096,0.051663110653559365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,3584,0.03903466794225905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,3072,0.05449777841567993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,3584,0.05938933293024699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,3072,0.03688977824317084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,3072,0.03993955585691664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,3584,0.045531554354561694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,2560,0.03353066576851739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,2560,0.05068622363938225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,2560,0.034061332543691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,2048,0.028032888968785603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,2048,0.029713776376512315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,1536,0.026850667264726426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,1536,0.042698668109046094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,1024,0.02204177776972453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,2048,0.046847111648983426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,1536,0.022337777747048274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,768,0.01242844429281023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,1024,0.03829244441456265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,768,0.020911999874644812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,512,0.009309333231714036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,1024,0.01571377780702379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,768,0.036425779263178505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,256,0.006507555643717448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,65536,0.4383466773562961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,512,0.01923199991385142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,512,0.03453777896033393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,128,0.0041519999504089355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,256,0.0158915552828047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,64,0.00388711111413108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,256,0.02959022257063124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,65536,0.48142576217651367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,128,0.015255111787054272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,128,0.027106665902667578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,32,0.003933333274390963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,64,0.01583377851380242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,32,0.016009777784347534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,16384,0.12641955746544734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,16384,0.10752355390124851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,12288,0.11658755938212077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,16384,0.1558133363723755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,12288,0.08322844240400526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,10240,0.07162755727767944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,12288,0.09883555438783433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,10240,0.09743289152781169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,10240,0.08741066853205363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,8192,0.06081777811050415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,8192,0.07937422063615587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,8192,0.07470933596293132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,7168,0.0544106662273407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,7168,0.07018222411473592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,65536,0.3707919915517171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,7168,0.06640088558197021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,6144,0.06077422036064995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,6144,0.04962311188379923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,6144,0.06557510958777533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,5120,0.04399111204677158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,5120,0.05048266715473599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,5120,0.05873066849178738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,4096,0.037624888949924044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,4096,0.05427289009094238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,3584,0.051058666573630444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,3584,0.034972445832358465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,4096,0.04188266727659437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,3072,0.033095998896492854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,3584,0.03691644469896952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,2560,0.030052443345387776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,2560,0.046486222081714206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,3072,0.04999822378158569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,2048,0.026555554734336004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,2048,0.04216622312863668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,65536,0.429604450861613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,3072,0.03159911102718777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,2048,0.023209777143266466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,1536,0.023200000325838726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,2560,0.027613333529896204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,1024,0.012735111018021902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,768,0.010776888993051318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,1024,0.021035555336210463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,1536,0.018070222602950204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,768,0.020033778415785897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,1024,0.03708444370163812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,1536,0.04203288753827413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,512,0.008332444561852349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,512,0.018093332648277283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,768,0.03571466604868571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,256,0.01609511176745097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,256,0.0053084443012873335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,512,0.03376177615589566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,128,0.003958222352796131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,256,0.028655999236636694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,128,0.015106666419241162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,64,0.0035777779089079965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,128,0.027432888746261597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,32,0.01579466627703773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,32,0.0038080000215106537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,64,0.01566755606068505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,65536,0.7635999785529243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,16384,0.09892800119188096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,16384,0.11054488685395981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,65536,0.5710053443908691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,12288,0.07697333229912652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,16384,0.1355839967727661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,12288,0.09176711241404216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,10240,0.06666755676269531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,12288,0.09109777874416775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,10240,0.08140089114507039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,10240,0.07775288820266724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,8192,0.05617066886689928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,8192,0.06267199913660686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,8192,0.06982844405704074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,7168,0.05530666642718845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,7168,0.06616177823808458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,7168,0.05511022276348538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,65536,0.33977511194017196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,6144,0.047021332714292735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,6144,0.062047998110453285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,5120,0.0417155557208591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,4096,0.051714665359920926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,5120,0.04110755523045858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,5120,0.056157330671946205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,4096,0.03604622350798713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,6144,0.05566755268308851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,4096,0.03396000133620368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,3584,0.033281776640150286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,3072,0.031191110610961914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,3584,0.048948443598217435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,3584,0.030072000291612413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,2560,0.028132445282406274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,3072,0.04816533459557427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,3072,0.026572444372706946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,2560,0.044079999128977455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,2048,0.019075555933846366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,2048,0.025024000141355727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,1536,0.014898666077189975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,1536,0.022463111413849726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,2048,0.04051555527581109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,1024,0.010878221856223213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,2560,0.022835556003782485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,1536,0.03911911116706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,1024,0.02048711147573259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,768,0.009179555707507664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,1024,0.03585066729121738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,512,0.007394666473070781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,768,0.019533332851197984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,512,0.01715288890732659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,768,0.03491200009981791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,256,0.004528000122971005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,512,0.030642665094799463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,128,0.003928889003064898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,256,0.015853333804342482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,128,0.014987554815080432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,64,0.0035742223262786865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,256,0.028394665982988145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,128,0.026667555173238117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,32,0.00360977773865064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,64,0.015442666080262927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,32,0.015569777952300178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,65536,0.3830000029669867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,16384,0.07694488763809204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,12288,0.060848891735076904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,16384,0.10407288869222005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,16384,0.10486755768458049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,12288,0.07238222493065728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,12288,0.08008978101942274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,10240,0.05326844586266411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,10240,0.07109422153896756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,10240,0.06126044193903605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,8192,0.049560887946022876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,65536,0.25036801232231987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,8192,0.04555022385385302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,8192,0.061850666999816895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,7168,0.04198844565285576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,7168,0.05849599838256836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,6144,0.03843644592497084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,7168,0.04398577743106418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,65536,0.3134844568040636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,6144,0.05555377735031975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,6144,0.03896977835231357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,5120,0.0344782206747267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,4096,0.030549334155188665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,65536,0.4919350941975911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,5120,0.032834665642844305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,3584,0.028382221857706707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,5120,0.050760888391070895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,3072,0.02661866611904568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,3584,0.04680888851483663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,4096,0.027124444643656414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,3072,0.04311733444531759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,2560,0.018432888719770644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,4096,0.04734311170048184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,2048,0.015298666225539314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,3584,0.024226667152510747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,2560,0.024711999628278945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,2560,0.04058311051792569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,1536,0.011969777444998423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,2048,0.02251111136542426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,3072,0.021365332934591506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,1536,0.020769778225156996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,2048,0.03849066628350152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,1024,0.009013333254390294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,1024,0.019487111104859244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,1536,0.03741866681310866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,1024,0.03533422284656101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,768,0.007714666426181793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,768,0.017656889226701524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,512,0.006271111054552927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,768,0.03414577907986111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,512,0.016369778248998854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,256,0.0042444442709287005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,512,0.029512001408471003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,128,0.014278221461508008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,64,0.0034773333205117118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,256,0.014861333701345654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,128,0.0036106668412685394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,256,0.02824266751607259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,128,0.027033777700530157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,32,0.0037066667444176147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,64,0.01460088955031501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,32,0.01478844384352366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,16384,0.07155999872419569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,16384,0.0919297801123725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,12288,0.06394044558207194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,12288,0.05675644344753689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,16384,0.09375911288791233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,12288,0.07572622431649102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,10240,0.05384177631802029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,10240,0.04961511161592272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,8192,0.04284266630808512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,10240,0.0681884421242608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,65536,0.39136356777615017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,65536,0.23102845085991752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,8192,0.04400711258252462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,8192,0.05921155876583523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,7168,0.03964089022742377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,7168,0.038896001047558255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,5120,0.03266222278277079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,6144,0.03631733192337884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,7168,0.059262222713894315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,6144,0.03455911080042521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,6144,0.05343288845486111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,5120,0.0488026671939426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,4096,0.02866666515668233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,5120,0.029622223642137315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,3584,0.026857778429985046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,4096,0.045888000064426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,65536,0.2937137815687391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,3584,0.04375288883845011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,4096,0.02429955535464817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,3072,0.025255999631351892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,3584,0.021683555510309007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,2560,0.016539555456903245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,3072,0.041823112302356295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,2560,0.023338667220539514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,2048,0.013536888692114087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,3072,0.019321777754359774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,2560,0.039650665389166936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,2048,0.02176977859603034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,1536,0.010964444114102257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,1536,0.020311110549502902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,2048,0.037645333343082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,1024,0.00847999999920527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,1536,0.03713422351413303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,1024,0.018990221950742934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,768,0.007329777710967594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,768,0.01646577815214793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,1024,0.03586755527390374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,512,0.005988444305128521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,768,0.032446222172843084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,512,0.01608888970481025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,256,0.003887999802827835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,256,0.01497244503762987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,512,0.02988622254795498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,128,0.003555555724435382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,256,0.028033776415718928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,128,0.014485332700941296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,64,0.003573333223660787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,128,0.026531555586391028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,64,0.014542222023010254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,32,0.003620444486538569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,32,0.014662222729788886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,16384,0.09228889147440593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,16384,0.06622666782803006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,12288,0.05575022432539198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,16384,0.08393066459231907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,65536,0.36313777499728733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,12288,0.05283733208974203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,10240,0.04682488905058967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,12288,0.07714044385486178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,10240,0.07370577918158637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,8192,0.0384204453892178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,8192,0.040545778142081365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,10240,0.04689777890841166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,8192,0.05987199809816149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,7168,0.05700088871849907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,7168,0.03735555542839898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,7168,0.03400889039039612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,6144,0.05373333228958977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,5120,0.03090488910675049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,6144,0.034613334470325045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,5120,0.04872088962131076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,4096,0.027277333868874445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,5120,0.02588444451491038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,4096,0.044962667756610446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,6144,0.030220445659425523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,3584,0.025813332862324182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,4096,0.02165422174665663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,3584,0.042768001556396484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,3072,0.02426044477356805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,65536,0.21233688460456002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,2560,0.01462044484085507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,3072,0.04072355561786228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,2560,0.022170666191313002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,3584,0.01943111088540819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,2048,0.012413333687517377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,2560,0.03927466604444716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,65536,0.27519822120666504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,2048,0.021295999487241108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,3072,0.016932444439993966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,1536,0.010103111465771994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,2048,0.03711466656790839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,1536,0.01982222166326311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,1024,0.008023999631404877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,1536,0.03672088848219977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,1024,0.01718933383623759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,1024,0.03487288951873779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,768,0.00703999979628457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,768,0.016541333662139047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,512,0.004871110949251386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,768,0.032884445455339216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,512,0.01586666703224182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,256,0.003858666867017746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,512,0.02923555506600274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,256,0.01477599971824222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,128,0.003489777859714296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,256,0.027767111857732136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,128,0.014304889572991265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,64,0.0032906666811969546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,128,0.02625422179698944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,64,0.014319999350441827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,32,0.003313777761326896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,32,0.014632888966136508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,16384,0.06154577599631416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,16384,0.08520266744825576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,16384,0.07042577531602648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,12288,0.04913333389494154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,12288,0.06881155570348103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,12288,0.05560266971588135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,10240,0.04359289010365804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,10240,0.04019733270009359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,10240,0.061887111928727895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,65536,0.29962044292026097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,65536,0.19506488906012642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,8192,0.03800800111558702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,7168,0.037552889850404524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,7168,0.05283911029497782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,8192,0.0554711123307546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,6144,0.03180355495876736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,8192,0.03276800115903219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,5120,0.028360890017615423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,6144,0.04998044504059685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,7168,0.029310223129060533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,5120,0.04600444436073303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,4096,0.025584888127115037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,6144,0.02613688839806451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,4096,0.04275555411974589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,5120,0.022218666142887537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,3584,0.01644888851377699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,3584,0.024459555745124817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,3072,0.014538667268223233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,3072,0.023198222120602924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,3584,0.04044444362322489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,2560,0.012381333443853589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,2560,0.02164711058139801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,3072,0.04014488723542955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,4096,0.018569777409235638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,2048,0.010590222146775989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,2560,0.03800444470511542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,2048,0.02106133268939124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,1536,0.008720888859695857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,2048,0.03626311156484816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,1536,0.01941688855489095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,1024,0.006888000087605582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,1536,0.036010666026009455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,65536,0.2721751001146105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,768,0.016153777639071148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,1024,0.01699555582470364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,768,0.006272888845867581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,1024,0.03401155604256524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,512,0.004239999999602635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,768,0.03093155556254917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,256,0.0037724446091387006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,512,0.01573155489232805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,512,0.02963199880388048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,256,0.01480888823668162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,128,0.0034213333080212274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,256,0.027789334456125896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,128,0.014157333307796054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,64,0.0032808888289663526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,32,0.0032755554550223877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,128,0.026166222161716882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,64,0.014431110686726041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,32,0.014631110760900708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,65536,0.24571734004550508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,16384,0.05804622173309326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,12288,0.046522667010625206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,16384,0.08259911007351346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,16384,0.05434133277999031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,12288,0.04353777898682488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,12288,0.06696533494525485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,8192,0.03614577651023865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,10240,0.04076177875200907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,10240,0.06045866674847073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,10240,0.04231999980078804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,8192,0.05292710993025038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,7168,0.033294220765431724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,7168,0.05006577902370029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,6144,0.030396444929970637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,8192,0.03170222375128005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,6144,0.04766399992836846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,5120,0.027671110298898485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,7168,0.02889066603448656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,6144,0.025900445050663416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,5120,0.0440906650490231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,65536,0.18594222598605684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,4096,0.0240666667620341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,5120,0.018900444110234577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,4096,0.041856888267729014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,3584,0.013744889034165276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,3584,0.02380888826317257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,3072,0.012364444633324942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,3072,0.022766222556432087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,3584,0.04030844569206238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,2560,0.010842666857772403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,4096,0.015560888581805758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,3072,0.03904177745183309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,2560,0.02130311065249973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,2560,0.03711288836267259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,2048,0.009301332963837517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,2048,0.03552177879545424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,2048,0.020454222957293194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,1536,0.007796444826655918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,1536,0.017818666166729398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,1536,0.03516888949606154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,1024,0.0063031112982167145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,1024,0.016588444511095684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,768,0.005113777601056629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,1024,0.032688889238569475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,768,0.0162444445821974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,65536,0.24491733974880645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,512,0.003914666672547658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,768,0.03012266755104065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,512,0.01549600064754486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,512,0.028583112690183852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,256,0.0035662220584021676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,256,0.014850666125615438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,64,0.0031066667288541794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,128,0.0032924444725116095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,256,0.027573333846198186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,128,0.014137778017255994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,128,0.02608088817861345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,32,0.003286222202910317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,64,0.014141332772043018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,32,0.014489778214030795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,16384,0.05626222160127428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,12288,0.04499288731151157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,16384,0.07979466517766316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,65536,0.1988604466120402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,12288,0.03898222247759501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,12288,0.06510666343900892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,10240,0.041117333703570895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,16384,0.052766223748524986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,8192,0.03457333313094245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,8192,0.05113955669932895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,10240,0.05867822302712334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,7168,0.031009776724709406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,8192,0.028789334826999243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,10240,0.033815999825795494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,6144,0.02921600143114726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,7168,0.0494959983560774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,65536,0.1748888889948527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,6144,0.046774221791161426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,5120,0.01695644524362352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,7168,0.022088888618681166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,5120,0.02681422233581543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,4096,0.01404355631934272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,5120,0.04359466830889384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,4096,0.024336889386177063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,6144,0.01978488928741879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,3584,0.012469333079126147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,4096,0.040799111127853394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,3584,0.02332088847955068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,3072,0.011200000014570026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,3584,0.03843999902407328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,65536,0.23028622733222115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,3072,0.021759111020300124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,2560,0.00999022192425198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,3072,0.03819200065400865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,2560,0.02104355560408698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,2048,0.008656889200210571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,2560,0.03712977634535895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,2048,0.0199644449684355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,1536,0.0075839997993575195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,2048,0.03547822104560004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,1536,0.01760088900725047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,1024,0.006114666660626729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,1536,0.03401599989997016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,1024,0.016623111234770883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,768,0.004751111070315043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,1024,0.030561778280470107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,768,0.016028443972269695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,512,0.0038382220599386427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,768,0.029998222986857098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,512,0.015544000599119397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,256,0.0034835553831524322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,512,0.02888622217708164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,65536,0.18133778042263457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,256,0.014555555250909595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,256,0.027845333019892376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,128,0.0032133333798911837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,128,0.01424088908566369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,64,0.003064888839920362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,128,0.026406221919589575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,64,0.014154665999942355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,32,0.003136000078585413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,32,0.01457244488928053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,12288,0.043394668234719165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,16384,0.07716711362202962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,16384,0.05419111251831055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,12288,0.034332444270451866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,10240,0.03863288958867391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,10240,0.057987557517157666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,16384,0.04626311196221245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,12288,0.06415288978152804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,8192,0.033399111694759794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,10240,0.02972266740269131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,8192,0.0510791109667884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,7168,0.031109333038330078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,8192,0.025090666280852422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,7168,0.04830311073197258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,65536,0.16948355568779838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,6144,0.02801333202256097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,5120,0.01787911189926995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,5120,0.025942222939597234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,6144,0.045978665351867676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,7168,0.022945778237448797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,4096,0.015630222029156156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,6144,0.020615110794703167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,5120,0.04235022266705831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,4096,0.023825777901543513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,3584,0.011454222102959951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,4096,0.04016000032424927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,3584,0.022797332869635686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,3072,0.01036622209681405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,3584,0.03812355465359158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,65536,0.2273040082719591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,3072,0.021723555194007024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,2560,0.0092622223827574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,2560,0.02068622244728936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,3072,0.03835466504096985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,2560,0.03690222236845229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,2048,0.008127999802430471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,2048,0.01904266741540697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,1536,0.007118222614129384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,2048,0.03535644544495477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,1536,0.017324444320466783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,1536,0.03428622086842855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,1024,0.00564444437623024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,1024,0.01644088824590047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,65536,0.15132266945309109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,1024,0.030935111973020766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,768,0.004282666577233208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,768,0.01588000026014116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,512,0.003999999827808804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,768,0.029882666137483384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,512,0.015354666444990369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,256,0.0035093331502543557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,512,0.029029332929187354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,256,0.014554666148291694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,128,0.003253333270549774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,256,0.0273991111252043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,128,0.014110222458839417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,64,0.003072000005178981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,128,0.025928000609079998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,32,0.003192000091075897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,64,0.01427288850148519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,32,0.014399111270904541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,16384,0.05128799875577291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,12288,0.04139822059207492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,16384,0.075345774491628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,12288,0.06299110915925768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,16384,0.036927110619015165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,10240,0.03612533211708069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,10240,0.057277335060967334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,8192,0.031706667608684964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,12288,0.029882666137483384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,8192,0.021846221552954778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,10240,0.026127111580636766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,8192,0.049885332584381104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,7168,0.028311110205120508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,7168,0.04790133237838745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,7168,0.020026667250527274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,65536,0.15763821866777208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,6144,0.01774666706720988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,6144,0.02722399930159251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,5120,0.013350221845838757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,6144,0.04470488760206434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,5120,0.025084444218211707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,4096,0.011306666665607028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,65536,0.21212622854444715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,4096,0.023193778263198003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,5120,0.0425182216697269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,3584,0.010479110810491772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,4096,0.03996710975964864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,3584,0.022284444835450914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,3072,0.009495110975371467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,3584,0.038324445486068726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,3072,0.021242666575643752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,65536,0.12963288360171848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,2560,0.008672888908121321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,3072,0.03797333439191183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,2560,0.019759999381171334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,2048,0.0075351107451650835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,2560,0.0370124446021186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,2048,0.018068444397714403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,1536,0.006688000013430913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,2048,0.035007112556033664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,1536,0.017024889588356018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,1024,0.005035555611054103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,1536,0.03251111176278856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,1024,0.016156444946924847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,768,0.004383999854326248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,1024,0.03081866767671373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,768,0.015607110328144498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,512,0.0038195554580953387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,768,0.029703110456466675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,512,0.015139554937680563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,256,0.0034604445099830627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,512,0.028809779220157202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,256,0.014631110760900708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,128,0.0031831111345026228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,256,0.027531555957264368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,128,0.014277334014574686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,64,0.0030417777597904205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,64,0.014038221703635322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,128,0.025964443882306416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,32,0.0030986666679382324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,32,0.014359111587206522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,16384,0.04930222365591261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,12288,0.03949422306484646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,16384,0.07178488704893324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,12288,0.02522666586769952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,12288,0.06080444653828939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,16384,0.03120888935195075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,10240,0.035042666726642184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,10240,0.05499200026194254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,8192,0.030636443032158747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,10240,0.022096888886557683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,8192,0.018228444788191054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,8192,0.04783733354674446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,7168,0.016955556141005624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,7168,0.029507554239696924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,7168,0.04534222351180183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,65536,0.14519288804796007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,6144,0.01550044450494978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,65536,0.20180088943905303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,6144,0.026667555173238117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,5120,0.014183999763594734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,6144,0.043819556633631386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,5120,0.023940443992614746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,4096,0.011891555454995898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,5120,0.04089333282576667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,4096,0.022660444180170696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,3584,0.01071200023094813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,4096,0.03920355439186096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,3584,0.022065778573354084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,3584,0.03835733400450812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,3072,0.009511111511124505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,3072,0.020554666717847187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,3072,0.03679644399219089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,2560,0.00795022232664956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,2560,0.019676445258988273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,2048,0.007135999699433644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,2560,0.035177777210871376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,2048,0.017733333839310538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,65536,0.1043599976433648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,2048,0.03251733382542928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,1536,0.006004444426960415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,1536,0.016708445217874315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,1536,0.032440887557135686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,1024,0.004467555632193883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,1024,0.016237333416938782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,768,0.004101333518822988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,1024,0.030588444736268785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,256,0.003439999909864532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,768,0.015635554989178974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,512,0.0037475555307335327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,768,0.029166221618652344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,512,0.015088889333936902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,512,0.02903111113442315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,128,0.0031759999692440033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,256,0.014546665880415173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,256,0.02723466687732273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,128,0.01406133340464698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,64,0.0029964444951878656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,128,0.02586311101913452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,64,0.013958222336239286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,32,0.003093333293994268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,32,0.014294221997261047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,16384,0.0459555553065406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,16384,0.025106666816605464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,12288,0.020231111182106864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,12288,0.03727644350793626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,16384,0.06893689102596708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,10240,0.017453332742055256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,12288,0.05689511034223768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,10240,0.032347556617524885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,8192,0.014736889137162102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,8192,0.029057777590221826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,10240,0.051721778180864125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,7168,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,8192,0.045455998844570585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,7168,0.027686221732033625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,6144,0.01262399968173769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,65536,0.14254400465223524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,6144,0.025380444195535447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,7168,0.043771555026372276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,5120,0.011530666715568967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,5120,0.02381244467364417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,4096,0.01055466632048289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,6144,0.04179555508825514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,4096,0.021890666749742296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,5120,0.040024889840020075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,3584,0.008848889006508721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,3584,0.02048711147573259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,3072,0.00813244448767768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,4096,0.03851911094453599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,3584,0.03593777616818746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,3072,0.019984000258975558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,2560,0.007662222617202335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,65536,0.1810071070988973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,3072,0.037068443165885076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,2560,0.018067555295096505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,2048,0.0069022224181228215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,2560,0.03341689043574863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,2048,0.017483555608325534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,2048,0.032178666856553816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,1536,0.005616000129116907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,1536,0.016699555847379897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,65536,0.08264533016416761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,1024,0.004412444515360726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,1536,0.031806220610936485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,1024,0.01607377827167511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,768,0.004061333421203825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,1024,0.030208888981077407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,768,0.015479111009173922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,512,0.003740444365474913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,768,0.029792000850041706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,512,0.014979556202888489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,256,0.003403555601835251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,512,0.028889778587553237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,256,0.014440000057220459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,128,0.003128000017669466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,256,0.027248889207839966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,128,0.013996443814701505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,128,0.02625244524743822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,64,0.0029208889851967492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,64,0.014058666096793281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,32,0.003067555526892344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,16384,0.01864266726705763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,32,0.01426488823360867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,16384,0.04420800010363261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,65536,0.1387288835313585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,16384,0.06649244493908353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,12288,0.015110222829712762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,65536,0.06190311246448093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,65536,0.17894400490654838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,12288,0.036068442795011736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,12288,0.05607822206285265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,10240,0.013648889131016202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,10240,0.03209422363175286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,8192,0.012114666402339935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,10240,0.05202755663130018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,8192,0.028468443287743464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,8192,0.04559288753403557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,7168,0.01127644462717904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,7168,0.027069333526823256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,7168,0.04381422201792399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,6144,0.012104888757069906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,6144,0.024860444996092055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,4096,0.02121155626244015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,6144,0.04192799992031521
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,5120,0.010695111420419482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,5120,0.02345955537425147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,4096,0.009354666703277165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,5120,0.039536890056398176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,3584,0.008902221918106079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,4096,0.03769422239727444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,3584,0.019509333703253005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,3584,0.03489333391189575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,3072,0.007970666719807519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,3072,0.01884711119863722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,2560,0.007269333634111617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,3072,0.0340666671593984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,2560,0.01796711153454251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,2560,0.03347377644644843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,2048,0.006199999815887875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,2048,0.01750666730933719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,2048,0.0319217774603102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,1536,0.0052533331844541765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,1536,0.01667555504375034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,768,0.029729776912265356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,1024,0.004583999928500918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,1536,0.03180088930659824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,1024,0.015936000479592215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,768,0.004084444294373195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,1024,0.029930667744742498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,768,0.015450666348139444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,512,0.003698666476541095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,512,0.014913777510325113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,256,0.0033395555284288195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,512,0.02870933214823405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,256,0.014343111051453484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,256,0.027073777384228174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,128,0.0030257778449190986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,128,0.013979555832015144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,64,0.0029235554652081597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,128,0.0262435558769438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,32,0.0029973333908451926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,64,0.013961777091026306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,32,0.014083556003040738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,16384,0.01536177761024899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,16384,0.0436462230152554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,12288,0.012753778033786349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,16384,0.06217955880694919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,65536,0.13711467054155138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,12288,0.035197334157096014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,10240,0.013958222336239286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,10240,0.03234488765398661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,12288,0.05407466822200351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,65536,0.16749333010779488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,8192,0.012304888831244575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,10240,0.049504889382256397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,8192,0.02844088938501146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,8192,0.044194665220048696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,7168,0.011559999651379056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,7168,0.02567111121283637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,65536,0.049676444795396596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,7168,0.04214222232500712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,6144,0.01148444414138794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,6144,0.024328889118300542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,5120,0.010063110954231685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,6144,0.04091111156675551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,5120,0.022281777527597215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,4096,0.00886044485701455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,5120,0.03729066583845351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,4096,0.02046133412255181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,4096,0.03503288825352987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,3584,0.00816888858874639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,3584,0.020184000333150227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,3072,0.007696888513035244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,3584,0.033442666133244835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,3072,0.018761777215533786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,3072,0.034056001239352755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,2560,0.006672888994216919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,2560,0.018146667215559218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,2048,0.005695111221737332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,2560,0.03275555703375075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,2048,0.017450667089886136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,2048,0.03143999973932902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,768,0.003946666502290302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,1536,0.004960000101062986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,1536,0.01683999929163191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,1536,0.03186310993300544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,1024,0.004300444490379757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,1024,0.016192888220151264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,1024,0.030052443345387776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,768,0.015460444821251763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,768,0.029324445459577773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,512,0.0035866668654812705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,512,0.015087111128701104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,512,0.02864711152182685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,256,0.0033111110743549135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,256,0.01456622282663981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,256,0.027129777603679236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,128,0.00305244450767835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,128,0.01407911048995124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,128,0.02573866645495097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,64,0.0029093333416514923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,64,0.014129777749379476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,32,0.0029297777348094513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,32,0.014408888088332282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,65536,0.037520001331965126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,16384,0.012493333054913415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,65536,0.13413688871595594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,65536,0.1500640047921075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,16384,0.04283022218280368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,12288,0.010750222537252637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,16384,0.05719821982913547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,12288,0.035381333695517644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,12288,0.049327999353408813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,10240,0.011242667006121742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,10240,0.03131466772821214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,10240,0.04591200086805555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,7168,0.03927288783921136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,8192,0.010278222461541494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,8192,0.028239111105600994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,8192,0.04195288817087809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,7168,0.009791110952695211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,7168,0.025615110993385315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,6144,0.010209778116808997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,6144,0.02407377792729272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,6144,0.03759288787841797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,5120,0.009586667021115622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,5120,0.02200977835390303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,5120,0.03478577733039856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,4096,0.008417777717113495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,4096,0.020350222786267597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,4096,0.0336480008231269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,3584,0.0079297779334916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,3584,0.019472888774342008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,3584,0.03257511059443156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,3072,0.007875555919276344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,3072,0.018557333283954196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,3072,0.03298133280542161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,2560,0.00620266670982043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,2560,0.017856889300876193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,2048,0.005597333527273602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,2560,0.03207644489076402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,2048,0.017280888226297166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,1536,0.004888888862397936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,2048,0.03113244308365716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,1536,0.016729777057965595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,1024,0.0042133331298828125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,1536,0.031325333648257785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,1024,0.01571466690964169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,768,0.003922666526503033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,1024,0.029708445072174072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,768,0.015440889530711703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,768,0.029100444581773546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,512,0.0035866668654812705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,512,0.014816888504558139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,512,0.028063111835055884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,256,0.0032177778581778207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,256,0.014334221680959066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,256,0.026908444033728704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,64,0.013804444836245643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,128,0.00300177786913183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,128,0.014228444960382251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,64,0.0028444443725877335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,128,0.02574577762020959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,32,0.002883555574549569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,16384,0.011403555671374003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,32,0.014119999276267158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,16384,0.04196000099182129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,65536,0.13162133428785536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,65536,0.14996711413065592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,12288,0.010070222119490305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,65536,0.02399377855989668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,16384,0.0565297802289327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,12288,0.033971554703182645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,10240,0.009150221943855286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,12288,0.048935112026002675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,10240,0.030121776792738173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,8192,0.00847822262181176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,10240,0.04527911212709215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,8192,0.026518222358491685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,7168,0.007910221815109253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,8192,0.038988444540235735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,7168,0.02532711128393809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,6144,0.007243555453088548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,7168,0.03874399926927354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,6144,0.023157333334287006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,5120,0.00740622232357661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,6144,0.03658488723966811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,5120,0.021677333447668288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,5120,0.03422844409942627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,4096,0.0069511110583941145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,4096,0.020404444800482858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,3584,0.00741511086622874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,3584,0.01921333372592926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,4096,0.03425688876046075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,3072,0.006984000404675801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,3584,0.032418668270111084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,3072,0.01828444500764211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,3072,0.03276800115903219
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,2560,0.006235555642180973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,2560,0.017519111434618633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,1536,0.016186666157510545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,2560,0.031606223848130964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,2048,0.005504000104135937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,2048,0.016961778203646343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,1536,0.0047973332305749255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,2048,0.03067822257677714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,1024,0.00416355538699362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,1536,0.031187554200490315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,1024,0.015420445137553744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,1024,0.029636445972654555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,768,0.0038693332009845306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,768,0.015103111664454142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,512,0.003495111233658261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,128,0.002992000016901228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,768,0.02870933214823405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,512,0.014653333359294467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,256,0.0031439999325407874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,64,0.013721778160995908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,512,0.028413332170910303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,256,0.014196443888876172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,256,0.02725155486000909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,128,0.013843555417325763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,128,0.025705777936511572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,64,0.002806222273243798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,32,0.0028204443968004654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,32,0.014151111245155334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,65536,0.015625778171751235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,65536,0.13032978110843235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,16384,0.008027555214034187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,65536,0.1497288942337036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,16384,0.041350222296184964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,12288,0.00793511089351442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,16384,0.05589244763056437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,12288,0.03331644336382548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,8192,0.026328000757429335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,10240,0.007332444190979004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,12288,0.047232889466815524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,10240,0.02988977895842658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,8192,0.007167999943097432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,10240,0.04337955514589945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,7168,0.0070017774899800616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,8192,0.038712888956069946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,7168,0.024714666936132643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,6144,0.00685066646999783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,7168,0.03801866703563266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,6144,0.02318222158484989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,5120,0.007251555720965068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,4096,0.020000888241661918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,6144,0.036474668317370944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,5120,0.021537777450349595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,4096,0.006832888970772426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,5120,0.03404088815053304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,3584,0.007280888656775157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,3072,0.03218311071395874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,4096,0.033546666304270424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,3584,0.019183999962276883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,3072,0.006973333656787872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,3584,0.03271999955177307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,3072,0.01834755473666721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,2560,0.006160000132189856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,2560,0.017525333497259352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,2560,0.03157066636615329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,2048,0.005477333234416113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,2048,0.01682933337158627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,2048,0.030976000759336684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,1536,0.004801777915822135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,1536,0.016196444630622864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,1024,0.0041546668443414895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,1536,0.03070666723781162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,1024,0.015611555841233997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,1024,0.029481778542200725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,768,0.0038133333954546186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,768,0.015083556373914083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,512,0.003495111233658261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,768,0.028723554478751287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,512,0.014635556273990206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,256,0.0031617778456873367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,512,0.028204444381925795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,256,0.014172444740931192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,64,0.013821333646774292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,256,0.02693066663212246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,128,0.002961777771512667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,128,0.01386222243309021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,64,0.002862222285734283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,128,0.025723555021815833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,32,0.002847111059559716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,32,0.013945778210957846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,65536,0.01316177762216992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,16384,0.007045333584149678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,65536,0.12975911299387613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,12288,0.006958222223652734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,16384,0.040030221144358315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,10240,0.007129777636792924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,12288,0.03262133399645487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,10240,0.02966044346491496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,8192,0.00703999979628457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,8192,0.026312000221676294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,7168,0.006799110935793982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,7168,0.024868445263968572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,6144,0.006760889043410619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,6144,0.02307555576165517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,5120,0.007104888558387756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,5120,0.02160888910293579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,4096,0.006696888970004187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,3584,0.007170666423108842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,4096,0.019911110401153564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,3584,0.019152889649073284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,3072,0.006887110984987683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,3072,0.018392889036072623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,2560,0.006129777679840724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,2560,0.01762488815519545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,2048,0.005478222337034013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,2048,0.016766221986876596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,1536,0.004798222333192825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,1536,0.016180444094869826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,512,0.01462222304609087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,1024,0.004117333226733738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,1024,0.015613334046469795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,768,0.003773333297835456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,768,0.01512977812025282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,512,0.0034622223012977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,256,0.0031146665828095544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,128,0.002921777880854077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,256,0.014157333307796054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,128,0.013769778112570444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,64,0.0028195555011431375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,64,0.013702222042613559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,32,0.002850666642189026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,32,0.014006222287813822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,65536,0.011553777588738335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,16384,0.007311999797821045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,10240,0.029409776131312054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,65536,0.1293075614505344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,16384,0.03988622294531928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,12288,0.006992888947327931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,12288,0.032637334532207914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,10240,0.007135999699433644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,8192,0.006980444822046492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,8192,0.02639466689692603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,7168,0.00685422205262714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,7168,0.024751999311976965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,6144,0.006717333363162146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,5120,0.007063111497296228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,6144,0.023188443647490606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,5120,0.021591999464564856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,4096,0.006654222392373615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,4096,0.019927110936906602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,3584,0.007134222322040134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,3584,0.019077334139082167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,3072,0.0069368887278768755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,3072,0.018262222409248352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,2560,0.006181333214044571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,2560,0.017630222770902846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,2048,0.005438222239414851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,1536,0.0047884442740016514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,2048,0.01676177812947167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,1536,0.01603555513752831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,1024,0.004108444270160464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,768,0.0038453332251972626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,1024,0.015531554818153381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,512,0.0034337778472238113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,768,0.015099555253982544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,512,0.01464355488618215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,256,0.003100444459252887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,256,0.014096000128322177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,128,0.002967111145456632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,128,0.01402933398882548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,64,0.002756444530354606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,64,0.013743110828929476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,32,0.0027831111931138565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,32,0.014000889327791003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,10240,0.24949600961473253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,12288,0.2971502145131429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,16384,0.3884773254394531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,10240,0.490463998582628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,12288,0.5935129059685601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,8192,0.2039839956495497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,8192,0.391574223836263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,7168,0.363797320259942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,7168,0.181715554661221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,16384,0.7685857878790961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,6144,0.15849955876668295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,6144,0.2964613437652588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,12288,0.35938310623168945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,5120,0.14059821764628092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,5120,0.24895644187927246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,16384,0.4327590995364719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,4096,0.11223467191060384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,4096,0.20199999544355604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,10240,0.30358844333224827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,3584,0.1007617778248257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,8192,0.25064267052544487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,3584,0.1781066656112671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,3072,0.15292621983422175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,3072,0.09178222550286187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,2560,0.07834133174684313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,2560,0.12945066557990179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,2048,0.10316266616185506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,7168,0.22350311279296875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,2048,0.0719662242465549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,1536,0.08119555314381917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,6144,0.1964391072591146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,1536,0.055083556307686694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,1024,0.05517244338989258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,1024,0.04517689016130236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,5120,0.1702453295389811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,768,0.04439466529422336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,768,0.04084622197681003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,3584,0.12575556172264948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,512,0.031446221801969744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,512,0.03684622380468581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,4096,0.14042754968007407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,256,0.018277333842383493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,256,0.03125422199567159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,3072,0.11761955420176189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,128,0.010484444598356882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,128,0.029125332832336426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,2560,0.09463200304243301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,64,0.007915555602974361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,64,0.028512001037597656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,32,0.006575111299753189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,32,0.026936888694763184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,1536,0.0749866697523329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,2048,0.08965777688556248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,256,0.03802311089303758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,512,0.04641155401865641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,128,0.03375377919938829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,1024,0.059567113717397056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,768,0.05271288752555847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,16384,0.11515822675493027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,12288,0.09093599849277073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,12288,0.14548356003231472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,16384,0.20559822188483345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,10240,0.12180088626013862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,10240,0.08763200044631958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,8192,0.0661111142900255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,65536,0.41198844379848903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,8192,0.1031893359290229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,65536,0.7553768687778049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,7168,0.06215377648671468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,7168,0.10501777463489109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,6144,0.07558844486872356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,6144,0.054135110643174916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,5120,0.06294488906860352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,5120,0.047981331745783486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,16384,0.14776799413892958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,12288,0.10801777574751113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,4096,0.04214577873547872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,4096,0.05110044280687968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,3584,0.04514666729503208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,10240,0.09679288996590509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,3584,0.03905066516664293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,8192,0.08245155546400282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,3072,0.036299556493759155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,3072,0.03942044576009115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,7168,0.08356177806854248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,2560,0.03358666764365302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,65536,0.4633910920884874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,2560,0.03322844372855292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,2048,0.02955289019478692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,2048,0.027648000253571406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,1536,0.024910221497217815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,6144,0.06989422109391955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,1536,0.02587377859486474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,1024,0.01587199999226464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,1024,0.022595556245909795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,5120,0.06443022357092963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,768,0.012063110868136087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,768,0.021195555726687115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,4096,0.057533330387539335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,512,0.009119110802809397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,512,0.01958577831586202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,3584,0.055923554632398814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,256,0.006279110908508301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,256,0.015807999504937064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,3072,0.05255377623769972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,128,0.004039999925427967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,128,0.015640000502268474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,2560,0.048723555273479886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,64,0.0036106668412685394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,64,0.016399999459584553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,32,0.003948444293604957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,32,0.0165928883685006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,2048,0.04472088813781738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,1536,0.0418835547235277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,1024,0.03762844536039565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,768,0.03527466787232293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,512,0.033166223102145724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,16384,0.14084000057644314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,16384,0.0940577785174052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,256,0.02782577938503689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,65536,0.3283768759833442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,12288,0.07537421915266249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,128,0.026910222238964502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,12288,0.10779555638631184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,10240,0.08986310826407538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,10240,0.0649813347392612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,8192,0.05591289202372233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,8192,0.07530933618545532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,7168,0.06415288978152804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,7168,0.05032711227734884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,65536,0.5808044539557563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,6144,0.04945422212282816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,6144,0.05633778042263455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,5120,0.04724444283379448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,5120,0.040369778871536255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,16384,0.11365421613057454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,4096,0.03878044419818454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,4096,0.03661600086424086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,12288,0.09218844440248276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,65536,0.4065839979383681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,3584,0.0343137780825297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,3584,0.03413510984844632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,10240,0.08206755585140653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,3072,0.030092444684770372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,3072,0.03188000122706095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,8192,0.07180088758468628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,2560,0.028933333026038274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,2560,0.025826666090223525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,2048,0.021350221501456365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,2048,0.025795555777019922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,7168,0.0682835578918457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,1536,0.01694222291310628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,1536,0.023560000790490046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,6144,0.06177688969506157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,1024,0.011774222056070963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,1024,0.021300445000330608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,5120,0.05753155549367269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,768,0.009673777553770278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,768,0.020184000333150227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,4096,0.05164444446563721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,512,0.0075502221783002215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,512,0.01794577803876665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,3584,0.050939556625154286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,256,0.0046968890560997855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,256,0.015319110618697273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,3072,0.04807910985416836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,128,0.0036951113078329298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,128,0.015368888775507608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,2560,0.04481600059403313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,64,0.003489777859714296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,64,0.015993777248594496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,32,0.003546666767862108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,32,0.01603200038274129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,2048,0.03999733262591892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,1536,0.03884889019860162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,1024,0.03569688730769687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,768,0.03476622369554307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,16384,0.08546933200624253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,16384,0.11652889516618516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,65536,0.29291733105977374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,12288,0.08887110816107856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,128,0.026405332816971674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,512,0.032984889215893216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,256,0.02748799986309475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,12288,0.06799555487102933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,10240,0.058931556012895375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,10240,0.07405333386527167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,8192,0.05048444535997179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,8192,0.0607137746281094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,7168,0.046810666720072426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,7168,0.05356088942951626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,65536,0.48437775505913633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,6144,0.047134220600128174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,6144,0.04589510957400004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,5120,0.03966133462058173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,5120,0.03804533349143134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,16384,0.10305600033866034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,4096,0.032829334338506065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,65536,0.35080888536241317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,4096,0.03461777832772996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,3584,0.029078221983379785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,12288,0.08337777853012085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,10240,0.07454399930106269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,3584,0.032249778509140015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,3072,0.029490666257010564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,3072,0.025590222742822435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,2560,0.022015111313925848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,8192,0.06636799706353082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,2560,0.02701866626739502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,2048,0.018171555466122098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,2048,0.024438222249348957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,7168,0.06275466415617201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,1536,0.014305777019924588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,1536,0.02299555473857456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,6144,0.057068447271982826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,1024,0.010341333018408881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,1024,0.02075822154680888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,5120,0.05377689003944397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,768,0.00868622213602066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,768,0.01962222158908844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,3584,0.04787910977999369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,4096,0.051233778397242226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,512,0.006834666762087081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,512,0.017264000243610807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,256,0.004138666722509596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,256,0.015132443772421943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,3072,0.04467111163669162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,128,0.003589333345492681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,128,0.015023999743991427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,2560,0.04207999838723076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,64,0.003403555601835251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,64,0.015535111228624979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,32,0.0035004446076022256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,32,0.015603555573357476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,2048,0.039896888865364924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,1536,0.03779200050565932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,1024,0.03547822104560004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,65536,0.2317404482099745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,768,0.033893333541022405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,16384,0.07021333111657037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,16384,0.09429777993096246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,512,0.030939555830425684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,12288,0.05708355373806424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,12288,0.07200977537367079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,256,0.02702933384312524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,128,0.026194666822751362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,10240,0.04997244477272034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,65536,0.38717155986362034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,8192,0.04299288988113403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,10240,0.06052355633841621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,8192,0.049504889382256397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,7168,0.0396364430586497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,7168,0.043735110097461276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,6144,0.03617422117127313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,6144,0.03862399856249491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,5120,0.032594667540656194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,5120,0.03268622358640035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,16384,0.09037688705656265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,4096,0.027152000202072993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,4096,0.028756442997190688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,65536,0.29702756139967174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,12288,0.07758844561047025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,3584,0.02681422233581543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,10240,0.06747021940019396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,3584,0.024002666274706524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,8192,0.06326933039559259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,3072,0.02121155626244015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,3072,0.02368800010946062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,2560,0.022272000710169475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,2560,0.018159111340840656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,2048,0.01495822270711263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,2048,0.020781333247820537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,7168,0.05709155400594076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,1536,0.011935999823941125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,1536,0.019309333629078336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,6144,0.05270311236381531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,1024,0.008945778012275696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,1024,0.01774666706720988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,5120,0.050125333997938365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,768,0.007624000310897827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,768,0.01495911180973053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,4096,0.04505777690145704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,512,0.00619555554456181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,3584,0.04394222299257914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,512,0.01442488862408532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,256,0.003929777691761653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,256,0.013590222431553734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,3072,0.041323555840386286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,128,0.003552000141806073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,128,0.013554666605260638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,2560,0.03902933332655165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,64,0.0033288889875014624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,64,0.014102222190962898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,32,0.0035457776652442086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,32,0.014362666341993542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,2048,0.036779556009504534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,1536,0.036328001154793635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,1024,0.03393866618474325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,16384,0.09070933527416652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,16384,0.06550488869349162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,65536,0.21417600578731963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,768,0.03327911098798116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,12288,0.053220444255405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,12288,0.06387022468778822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,10240,0.054487999942567616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,10240,0.04650400082270304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,512,0.02848177817132738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,256,0.026685333914226953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,128,0.026168000366952684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,8192,0.04036622246106466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,8192,0.04557866520351834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,7168,0.03707733419206407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,65536,0.3544604513380263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,7168,0.038329776790406965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,6144,0.03423644436730279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,6144,0.034415112601386175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,5120,0.03308355477121141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,5120,0.030702223380406696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,16384,0.08552444643444485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,4096,0.025238222546047632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,65536,0.2744657728407118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,4096,0.0266248881816864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,12288,0.07176978058285184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,3584,0.02202933364444309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,3584,0.024832889437675476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,3072,0.01947111056910621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,3072,0.022894221875402663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,10240,0.06430577569537692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,2560,0.01683644453684489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,8192,0.05722577704323662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,2560,0.021420444051424663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,2048,0.013940445250935025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,2048,0.02017511096265581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,7168,0.05520088805092705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,1536,0.010819555156760745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,1536,0.018917333748605516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,6144,0.050796445873048574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,1024,0.008676444490750631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,1024,0.017440888616773818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,5120,0.04873422119352552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,768,0.007224000162548489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,768,0.014696889453464083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,4096,0.042945779032177396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,512,0.005391110976537068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,512,0.014197332991494073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,3584,0.04277511106597053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,256,0.00388711111413108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,256,0.01349955548842748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,3072,0.04071022073427836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,128,0.003567111161020067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,128,0.01346666696998808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,2560,0.03831377625465393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,64,0.0032968889507982465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,64,0.01382844481203291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,32,0.0033662221911880705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,2048,0.03594844539960226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,32,0.013805333111021252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,1536,0.035982221364974976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,1024,0.03415911065207587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,768,0.032856888241238065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,65536,0.19663466347588432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,16384,0.059637334611680776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,16384,0.07882489098442926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,12288,0.04904266529613071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,12288,0.06176800198025174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,256,0.02662044432428148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,512,0.028736889362335205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,65536,0.2921164565616184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,10240,0.052298665046691895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,128,0.026276444395383198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,10240,0.04669955703947279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,8192,0.0376835564772288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,8192,0.0398248897658454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,7168,0.03442666596836514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,7168,0.03509333398607042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,6144,0.031743112537595965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,6144,0.030799110730489094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,5120,0.025739555557568867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,5120,0.028544889556037054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,16384,0.07919110854466756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,65536,0.2334053251478407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,4096,0.024023110667864483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,12288,0.06614933411280315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,4096,0.025230222278171118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,3584,0.0188711103465822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,3584,0.023443554838498432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,3072,0.016747555798954435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,10240,0.06002666552861532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,3072,0.02167466613981459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,8192,0.05420177843835619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,2560,0.01440622243616316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,2560,0.020849777592553034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,2048,0.012079999678664737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,2048,0.019621334142155118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,7168,0.05471555391947428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,1536,0.009927999642160203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,1536,0.0184542222155465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,6144,0.047916445467207164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,1024,0.007827555967701806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,1024,0.015505777464972602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,5120,0.04626311196221245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,768,0.006832888970772426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,768,0.014558222558763293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,4096,0.04130222068892585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,512,0.00489688871635331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,512,0.014148443937301636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,3584,0.04046488801638285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,256,0.003710222327046924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,256,0.01351199961370892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,3072,0.0389048887623681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,128,0.0033359999457995095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,128,0.013374221821626028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,2560,0.03713688916630215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,64,0.0032417778339650896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,64,0.013630222115251752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,2048,0.03512888815667894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,32,0.0032319999817344877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,32,0.01388177772363027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,1536,0.03613244493802389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,1024,0.03322133421897888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,768,0.03122933374510871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,16384,0.06590755780537923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,16384,0.05390755666626824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,65536,0.16737243864271376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,12288,0.05173244410090976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,12288,0.044405334525638156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,10240,0.03969688879119025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,65536,0.24294755193922254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,10240,0.039388444688585066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,8192,0.032572444942262434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,512,0.028301331732008193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,256,0.026749332745869953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,8192,0.034520887666278414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,7168,0.03191377719243368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,7168,0.028675556182861328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,128,0.025941333836979334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,6144,0.02903111113442315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,6144,0.025849777791235182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,5120,0.022099556194411382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,5120,0.026195555925369263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,16384,0.0744088888168335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,65536,0.20999644862280953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,4096,0.018425777554512024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,4096,0.02296888828277588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,3584,0.01643466618325975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,3584,0.021728888154029846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,12288,0.062288886970943876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,3072,0.01443911095460256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,3072,0.020602666669421725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,10240,0.05686222182379829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,2560,0.014106666048367819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,2560,0.01958044370015462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,8192,0.051552888419893056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,2048,0.010544000400437249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,2048,0.018852444158660043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,7168,0.04962133367856344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,1536,0.008901333643330468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,1536,0.017907554904619854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,6144,0.045893334680133395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,1024,0.007027555671003129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,1024,0.014701333310869006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,5120,0.04413333204057482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,768,0.006284444282452266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,768,0.014489778214030795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,4096,0.040031999349594116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,512,0.004238222208287981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,512,0.013991999957296582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,3584,0.03934577769703335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,256,0.003660444584157732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,256,0.013429332938459186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,3072,0.03817955652872721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,128,0.003350222276316749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,128,0.013560888667901358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,2560,0.03641244437959459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,64,0.003134222287270758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,64,0.013614222407341003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,32,0.003145777723855443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,32,0.013430222041077085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,2048,0.03447022371821933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,1536,0.03505333264668783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,1024,0.033139556646347046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,768,0.030232889784706965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,16384,0.057479110029008657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,65536,0.1461146672566732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,16384,0.04821777674886915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,65536,0.19532977210150826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,12288,0.04243555665016174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,12288,0.04038222299681769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,10240,0.035848889085981585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,10240,0.036438223388459944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,8192,0.030639999442630347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,512,0.028219554159376357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,256,0.02664533257484436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,8192,0.03129155437151591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,128,0.026287110315428838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,7168,0.028973334365420874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,7168,0.027785778045654297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,6144,0.026073777013354834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,6144,0.02181511123975118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,5120,0.018756444255510967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,5120,0.023803555303149756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,16384,0.06871022118462457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,4096,0.015344889627562629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,4096,0.02158222264713711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,12288,0.059486223591698535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,3584,0.013633777697881063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,3584,0.02052533295419481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,65536,0.18243911531236437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,3072,0.012311110893885294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,3072,0.019831111033757527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,10240,0.05235822333229912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,2560,0.010833777487277985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,2560,0.018977777825461496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,8192,0.04741155438952976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,2048,0.009287999735938178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,2048,0.017975111802419026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,7168,0.04605866803063286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,1536,0.007897777689827813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,1536,0.015895111693276297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,6144,0.04221333397759331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,1024,0.00628977765639623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,1024,0.014533332652515836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,5120,0.04109155469470554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,768,0.004987555659479565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,768,0.014255111416180929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,4096,0.037680887513690524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,512,0.003958222352796131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,512,0.014059555199411182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,3584,0.037583112716674805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,256,0.0034355556385384668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,256,0.013371555341614617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,3072,0.0369395547442966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,128,0.0032097777972618737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,128,0.013342222405804528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,2560,0.03553244471549988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,64,0.002990222225586573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,64,0.013598221871587964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,32,0.00305600009030766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,32,0.013375110924243927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,2048,0.03403110967742072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,1536,0.03473511007097032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,1024,0.03032711148262024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,768,0.02895555562443203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,16384,0.04560710986455282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,16384,0.051162666744656034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,65536,0.13717155986362034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,12288,0.038057777616712786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,12288,0.0398248897658454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,65536,0.17036799589792886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,10240,0.03418666786617703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,10240,0.033949332104788885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,8192,0.03179999854829576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,8192,0.02938488788074917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,512,0.028129776318868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,7168,0.02570666703912947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,256,0.02643999954064687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,7168,0.027611555324660406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,128,0.025795555777019922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,6144,0.023258666197458904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,6144,0.025586666332350835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,5120,0.016691555579503376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,5120,0.022595556245909795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,16384,0.06534400251176622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,4096,0.013716444373130798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,4096,0.020962667134073045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,65536,0.16913866996765137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,3584,0.012483555409643384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,12288,0.05498399999406603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,3584,0.020152888364262052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,3072,0.011357333097192975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,3072,0.019389332996474374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,10240,0.05052711235152351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,2560,0.009951111343171861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,2560,0.018699554933442008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,8192,0.045941332976023354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,2048,0.009000889129108852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,2048,0.017582222819328308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,7168,0.04407466782463921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,1536,0.007510222494602203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,6144,0.041195554865731135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,1536,0.015230221880806817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,5120,0.0395448903242747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,1024,0.006105777704053455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,1024,0.014431110686726041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,768,0.004398222184843487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,768,0.014343111051453484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,4096,0.03714044557677375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,512,0.00388355553150177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,512,0.01385244478782018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,3584,0.03705777724583944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,256,0.0034248888906505373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,256,0.01333333303531011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,3072,0.03659644391801622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,128,0.0032151111712058387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,128,0.01348355578051673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,2560,0.03504622313711379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,64,0.003039999968475766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,64,0.013415999710559845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,32,0.003035555490189128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,32,0.013222222526868185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,2048,0.033999999364217125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,65536,0.13226221667395696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,1536,0.03369777732425266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,65536,0.15078311496310765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,16384,0.04374044471316867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,1024,0.030072887738545735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,16384,0.04264533188607958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,768,0.02903199858135647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,12288,0.036248889234330915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,12288,0.03373688790533278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,10240,0.03224622209866842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,10240,0.029636445972654555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,8192,0.024634665913052026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,8192,0.02805866797765096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,512,0.027914666467242773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,7168,0.021873777111371357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,7168,0.02676711148685879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,256,0.026438222991095647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,128,0.026211554805437725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,6144,0.01755644381046295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,6144,0.02372444503837162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,5120,0.014870221416155497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,5120,0.02217777735657162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,16384,0.06234221988254123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,4096,0.012292444705963135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,4096,0.02053244411945343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,12288,0.053298665417565234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,3584,0.011424000064531961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,3584,0.01997777819633484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,10240,0.04860088891453213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,3072,0.010291555689440833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,3072,0.01904622217019399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,65536,0.15951289070977107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,2560,0.009207111265924241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,2560,0.018439110782411363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,8192,0.04400888747639126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,2048,0.008080888953473832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,2048,0.01706044375896454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,7168,0.04279911186960009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,1536,0.007064888874689738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,1536,0.01512888901763492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,6144,0.03990933299064636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,1024,0.005345777918895085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,5120,0.03965866565704346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,1024,0.014369777507252164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,768,0.004327110946178436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,768,0.014075555735164218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,4096,0.03611200054486593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,512,0.003908444609906939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,512,0.013983999689420065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,3584,0.03674755493799845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,256,0.0034737777378824023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,3072,0.036107556687461004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,256,0.013400000002649097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,128,0.00319377767542998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,128,0.013311111264758639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,2560,0.03487911158137851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,64,0.003075555587808291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,64,0.013463111387358772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,32,0.003084444544381566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,32,0.013367999758985309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,2048,0.033168001307381526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,65536,0.12780355082617864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,1536,0.0325911111301846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,1024,0.030241777499516804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,768,0.028637333048714533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,16384,0.04107644491725498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,16384,0.03613777624236213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,12288,0.0350542234049903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,65536,0.1280062198638916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,12288,0.02914666798379686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,10240,0.03018488817744785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,10240,0.025295999315049913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,8192,0.02148444453875224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,8192,0.027003554834259882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,128,0.025785777303907607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,256,0.026334222820070054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,7168,0.017256889078352187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,512,0.02794399857521057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,7168,0.025044444534513686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,6144,0.015434667468070984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,6144,0.023256889647907678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,5120,0.013208889298968844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,5120,0.021665778425004747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,16384,0.06002666552861532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,4096,0.011156444748242697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,4096,0.02010577751530541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,65536,0.15149511231316462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,3584,0.010231999887360467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,3584,0.019425777925385367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,12288,0.050711999336878456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,3072,0.009431111315886179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,3072,0.018975999620225694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,10240,0.047154668304655284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,2560,0.008565333154466417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,2560,0.017700443665186565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,8192,0.04274133178922865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,2048,0.007490666376219855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,2048,0.01572177807490031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,7168,0.04199999901983473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,1536,0.006604444649484422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,1536,0.015160888433456421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,6144,0.038754665189319186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,1024,0.004961777892377642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,1024,0.014519999424616495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,5120,0.03851911094453599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,768,0.004200889004601372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,768,0.014143110977278816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,4096,0.03568444318241543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,512,0.0037671112351947357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,3584,0.036307556761635676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,512,0.01385155568520228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,256,0.0034053333931499054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,256,0.013276444541083442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,3072,0.03579555617438422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,128,0.003160888950030009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,128,0.013304000099500021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,2560,0.03503466645876566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,64,0.002999111182159848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,64,0.01347733371787601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,32,0.003068444422549672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,32,0.013443555268976422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,2048,0.03251733382542928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,65536,0.12273422876993816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,1536,0.031242665317323472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,65536,0.10548533333672418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,16384,0.030047999487982854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,16384,0.03873511155446371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,12288,0.024851555625597637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,12288,0.03173689047495524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,1024,0.029768887493345473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,10240,0.021385777327749465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,10240,0.029537777105967205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,768,0.028824888997607764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,8192,0.018043556147151522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,8192,0.025975111458036635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,256,0.026289777623282537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,512,0.02794044547610813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,7168,0.016380444169044495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,7168,0.024900444679790076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,6144,0.01572533282968733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,6144,0.02277511192692651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,128,0.02566755480236477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,5120,0.013239999612172445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,5120,0.02149244480662876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,16384,0.05808622307247586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,4096,0.011724444727102915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,4096,0.02004888819323646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,65536,0.14434311125013563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,3584,0.009883555273214975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,12288,0.048991110589769155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,3584,0.019336889187494915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,3072,0.008544888761308458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,3072,0.018214222457673814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,10240,0.04482577906714546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,2560,0.007841777470376756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,2560,0.01609155535697937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,8192,0.04205066627926297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,2048,0.007063111497296228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,2048,0.015462223026487561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,7168,0.040103111002180315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,1536,0.005828444328573015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,1536,0.014720888601409065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,6144,0.03807377815246582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,1024,0.004351110921965705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,1024,0.014333332578341166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,5120,0.038089778688218855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,768,0.004108444270160464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,768,0.014337778091430664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,4096,0.03573155403137207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,512,0.003685333248641756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,512,0.014031110538376702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,3584,0.035996443695492215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,256,0.003335111050142182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,256,0.013519110778967539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,3072,0.0351857774787479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,128,0.003125333330697484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,128,0.013350221845838757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,2560,0.03346133232116699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,64,0.0029519999192820657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,2048,0.030249777767393324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,64,0.013163555827405719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,32,0.002974222310715251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,32,0.012989333106411828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,1536,0.03087644444571601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,1024,0.029647111892700195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,65536,0.07996888955434163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,16384,0.02477955487039354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,65536,0.11571378178066677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,16384,0.037688887781567044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,12288,0.03144888745413886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,12288,0.019910222954220243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,10240,0.01737511157989502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,10240,0.027902222341961328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,768,0.02881244487232632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,8192,0.014708444476127625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,8192,0.0251920007997089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,512,0.02790844440460205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,7168,0.013567999833159976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,7168,0.023997333314683702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,256,0.026667555173238117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,6144,0.01276000009642707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,6144,0.021744888689782884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,128,0.026110221942265827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,5120,0.01183733344078064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,5120,0.021006221572558086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,12288,0.04673777686225044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,16384,0.05537244346406725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,4096,0.01051288925939136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,4096,0.018903111418088276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,65536,0.1404257747862074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,3584,0.008763555851247575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,3584,0.017722666263580322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,3072,0.008159110943476358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,10240,0.044248888889948525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,3072,0.016295111841625638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,2560,0.007490666376219855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,2560,0.015776000089115567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,8192,0.04046400056944953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,2048,0.0063075555695427795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,2048,0.015266666809717814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,7168,0.0399600002500746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,1536,0.0052871112194326185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,1536,0.014801777071423002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,6144,0.038345777326160006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,1024,0.00433688900536961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,1024,0.014308444327778287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,5120,0.03744622071584066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,768,0.00397155558069547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,768,0.014075555735164218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,4096,0.03532888823085361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,512,0.00361333332127995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,3584,0.03396711084577773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,512,0.013780444032616086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,256,0.003318222239613533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,256,0.013304000099500021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,3072,0.03227644496493869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,128,0.003044444446762403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,2560,0.03151022063361274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,128,0.01313777764638265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,64,0.0028968888024489083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,64,0.01313866674900055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,32,0.0029057777590221832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,32,0.012940444052219391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,2048,0.02988977895842658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,65536,0.055098666085137256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,65536,0.08485955662197536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,16384,0.018353778454992507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,16384,0.029097778929604426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,1536,0.030697776211632624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,12288,0.014666666587193808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,12288,0.025597333908081055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,1024,0.030132446024152968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,10240,0.013273778061072031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,768,0.029399110211266413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,10240,0.02360711163944668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,8192,0.011606222225560082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,8192,0.02148799929353926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,256,0.026105778084860906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,7168,0.011150221857759686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,512,0.027836445305082533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,7168,0.020578665865792167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,128,0.02570666703912947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,6144,0.011121778024567498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,6144,0.019660444723235238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,5120,0.010248888697889116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,5120,0.01757866640885671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,65536,0.13644799921247694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,4096,0.009354666703277165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,4096,0.016332444217469957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,16384,0.05343022280269199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,3584,0.009062222308582729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,3584,0.01625333395269182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,12288,0.04633333285649618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,3072,0.008292444050312042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,3072,0.01575199928548601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,10240,0.04320177767011854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,2560,0.006839111033413146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,2560,0.015250666273964776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,8192,0.0403457780679067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,2048,0.005657777604129579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,2048,0.014967999524540372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,7168,0.03947110970815023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,1536,0.004942222187916438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,1536,0.014602666099866232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,6144,0.037422223223580256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,1024,0.004275555411974589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,1024,0.014004444082578024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,5120,0.03703644540574815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,768,0.003960000144110786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,768,0.013830222189426422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,4096,0.034019556310441755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,512,0.0035822221802340615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,512,0.013603555659453074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,3584,0.03312177790535821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,256,0.0032542221662071017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,256,0.013199110825856527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,3072,0.032305777072906494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,128,0.003040888864133093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,128,0.013032889200581444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,2560,0.031014220582114324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,64,0.002945777856641345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,64,0.013032000097963544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,32,0.00301155551440186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,32,0.012715555727481842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,65536,0.04782311121622721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,2048,0.030020445585250854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,65536,0.08065244224336413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,16384,0.015291555060280694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,16384,0.028408888313505385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,1536,0.030656887425316706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,1024,0.029492444462246362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,12288,0.012891555825869242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,12288,0.02490844494766659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,10240,0.012025777664449481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,768,0.02847911251915826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,10240,0.02313066687848833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,8192,0.01075644459989336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,8192,0.021129777034123737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,512,0.027645332945717707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,7168,0.010422222316265106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,7168,0.020091555184788175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,256,0.026188444760110643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,6144,0.009639110830095079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,6144,0.01850488947497474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,128,0.025711110896534387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,5120,0.009136888715955947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,5120,0.01756800048881107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,16384,0.05273155702484978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,4096,0.007854222423500484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,4096,0.01631644368171692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,65536,0.13632710774739584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,3584,0.008307555483447181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,3584,0.016021332807011075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,12288,0.0461351109875573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,3072,0.007631999750932057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,3072,0.015561777684423657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,10240,0.04329599936803182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,8192,0.03979911075698005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,2560,0.0064080001579390625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,2560,0.01513066722287072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,2048,0.005597333527273602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,2048,0.014697778556081982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,7168,0.03944799966282315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,1536,0.00488800017370118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,1536,0.014270222849316068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,6144,0.036325332191255354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,1024,0.004238222208287981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,1024,0.014196443888876172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,5120,0.03645688957638211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,768,0.003940444439649582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,768,0.014013333453072442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,4096,0.032482667101754084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,512,0.003617777592606015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,512,0.013624889155228933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,3584,0.03288799855444167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,256,0.0032524443748924467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,256,0.013316444224781461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,3072,0.03186310993300544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,128,0.0030008889734745026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,128,0.012960000170601739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,2560,0.03124089042345683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,64,0.0028737777223189673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,64,0.013020444247457715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,32,0.0028782222006056043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,32,0.01273333364062839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,2048,0.02959377898110284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,65536,0.03568266828854879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,65536,0.0773564444647895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,16384,0.012061333490742577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,16384,0.02805244591501024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,1536,0.030441777573691473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,12288,0.010751999914646149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,12288,0.0244835548930698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,1024,0.029637333419587877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,10240,0.010002666877375709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,10240,0.022311111291249592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,768,0.02851822310023838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,8192,0.009269333548016017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,8192,0.019859555694792006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,512,0.027818666564093694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,7168,0.008851555486520132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,7168,0.018961777289708454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,256,0.02641244398223029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,6144,0.008408000071843466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,6144,0.017811555001470778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,128,0.025711999999152288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,5120,0.008112000094519721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,5120,0.01700888905260298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,16384,0.05272622240914238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,4096,0.007333333293596904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,65536,0.13646311230129665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,4096,0.016135111451148987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,3584,0.007687110867765214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,3584,0.015710221396552194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,12288,0.0461297763718499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,3072,0.007224889265166388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,3072,0.015245333313941956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,10240,0.0432204438580407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,2560,0.00630666646692488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,2560,0.014988443917698331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,8192,0.039790223042170204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,2048,0.005579555614127054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,2048,0.014700444208251106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,7168,0.039164443810780845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,1536,0.004888888862397936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,1536,0.01427111029624939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,6144,0.03562488820817735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,1024,0.004237333519591225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,1024,0.014038221703635322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,5120,0.03436888919936286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,768,0.0038071109188927543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,768,0.013758222262064615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,4096,0.03222577770551046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,3584,0.03258133265707228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,512,0.003528888854715559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,512,0.013456000222100152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,256,0.003176888864901331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,256,0.013078221844302284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,3072,0.03175110949410333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,128,0.0029688889367712867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,128,0.012999110751681857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,2560,0.030945777893066406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,64,0.0028693332440323303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,64,0.012913777596420713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,32,0.0028595555987623003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,32,0.012617777619096967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,65536,0.023260444402694702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,65536,0.07470399803585477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,2048,0.029559999704360962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,16384,0.009293333523803288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,16384,0.02791555391417609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,1536,0.03050755461057027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,12288,0.008201777935028076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,12288,0.024241778585645888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,1024,0.029670221938027277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,10240,0.008192888564533657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,10240,0.021406221720907424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,768,0.02823466724819607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,8192,0.0074675555030504865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,8192,0.01941422290272183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,512,0.027627555860413447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,7168,0.007195555501514011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,7168,0.018602665927675035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,256,0.025994666748576697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,6144,0.0069475554757648045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,6144,0.01761244402991401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,128,0.025654221574465435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,5120,0.007320889168315464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,5120,0.01681600014368693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,65536,0.13720088534884983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,4096,0.0068862222962909276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,4096,0.016234666109085083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,16384,0.0529075562953949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,3584,0.00740622232357661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,3584,0.01586577792962392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,12288,0.04562933246294657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,3072,0.0069982219073507525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,3072,0.015264888604482016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,10240,0.04124800033039517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,2560,0.006156444549560547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,2560,0.015000000596046448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,8192,0.03884622123506334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,2048,0.005472888963090048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,2048,0.014536889062987434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,7168,0.03706666827201843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,1536,0.004820444517665439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,1536,0.014177777700954013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,6144,0.03464355402522617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,1024,0.0041680000722408295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,1024,0.013775111072593264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,5120,0.03435111045837402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,768,0.0038044444388813446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,768,0.013651555611027611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,4096,0.03198577629195319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,512,0.003519999898142285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,512,0.013479111095269522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,3584,0.03264800045225356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,256,0.0031840000301599503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,256,0.013183111117945777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,3072,0.03198577629195319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,128,0.002999111182159848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,128,0.012989333106411828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,2560,0.03107200066248576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,64,0.0028693332440323303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,64,0.01295733369059033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,32,0.0028444443725877335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,32,0.012773333324326409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,65536,0.01442577772670322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,65536,0.07568444146050347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,2048,0.029911110798517864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,16384,0.00720977783203125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,16384,0.02648977769745721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,1536,0.030100444952646892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,12288,0.006903111106819577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,12288,0.022826666633288067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,1024,0.029206222958034937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,10240,0.007149332927332983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,10240,0.02109066645304362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,768,0.02804888950453864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,8192,0.007102222078376346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,8192,0.01923022170861562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,512,0.027479999595218237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,7168,0.0069466663731469055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,7168,0.018588445252842374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,256,0.026367111338509455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,6144,0.006748444504208035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,128,0.02584444483121236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,6144,0.017646221650971305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,5120,0.0071937781241205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,5120,0.01686755485004849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,65536,0.1371928850809733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,4096,0.006726222319735422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,4096,0.016200888488027785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,16384,0.05036088824272156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,3584,0.007157333195209503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,3584,0.01571911076704661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,12288,0.04334222277005514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,3072,0.006861333217885759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,3072,0.015311999453438653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,10240,0.0403511126836141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,2560,0.006205333189831839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,2560,0.014852444330851236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,8192,0.03693155447642008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,2048,0.005471999860472149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,2048,0.014444443914625378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,7168,0.03674222363366021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,1536,0.004824000100294749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,1536,0.014330666926172046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,6144,0.03469333383772109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,1024,0.004182222402758068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,1024,0.01399466726515028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,5120,0.034408890538745456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,768,0.003798222376240624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,768,0.013557333085272046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,4096,0.03238577644030253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,512,0.003510222252872255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,512,0.01370044466522005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,3584,0.03287555442916022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,256,0.003117333269781537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,256,0.013136000268989138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,3072,0.03195644418398539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,128,0.0029768887907266617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,128,0.013048888908492194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,2560,0.030814223819308813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,64,0.0028115556471877625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,64,0.012805333567990197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,32,0.00282577777074443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,32,0.012498666842778524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,65536,0.013560888667901358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,65536,0.07504266500473022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,16384,0.007153777612580194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,16384,0.025811556312772963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,12288,0.006855999843941794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,12288,0.022524444593323603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,10240,0.007157333195209503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,2048,0.02972444560792711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,10240,0.02108000053299798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,8192,0.006950221955776215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,8192,0.019276445110638935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,7168,0.006744888921578725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,6144,0.006605333338181178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,7168,0.01847022275129954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,5120,0.006986666884687211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,6144,0.017487110363112558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,5120,0.016742222838931613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,4096,0.00655644428398874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,4096,0.016065778003798593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,3584,0.0069955554273393415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,3072,0.006710222197903528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,3584,0.015631111131774057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,3072,0.015106666419241162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,2560,0.006150222072998683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,2560,0.014701333310869006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,2048,0.005479111025730769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,1536,0.0047973332305749255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,2048,0.014418666561444601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,1024,0.004117333226733738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,1536,0.014071111877759298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,768,0.0037946667936113146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,1024,0.013728888498412238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,768,0.013564444250530668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,512,0.0034373334298531213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,512,0.01330488920211792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,256,0.003120888852410846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,256,0.012990222209029727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,128,0.002973333415057924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,128,0.012847111456924014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,64,0.0027884443600972495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,64,0.01276266657643848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,32,0.003362666608558761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,32,0.01240711079703437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,65536,0.012206222448084088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,65536,0.0746684471766154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,16384,0.006911110960774952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,16384,0.025759999950726826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,12288,0.006831999868154526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,1536,0.030191110240088567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,10240,0.006968888971540663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,12288,0.022406222091780767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,8192,0.006837333242098491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,10240,0.02103022237618764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,7168,0.006705777926577463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,8192,0.01921866668595208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,6144,0.006513777706358168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,7168,0.01865955524974399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,5120,0.0069164443347189165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,6144,0.01754133403301239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,4096,0.00646577775478363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,5120,0.01696266730626424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,4096,0.015982222225930955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,3584,0.006877333339717653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,3584,0.015614221493403116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,3072,0.006497777584526274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,3072,0.01518133282661438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,2560,0.0061039999127388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,2048,0.005440000030729506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,2560,0.014671999547216626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,2048,0.014470222923490735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,1536,0.004736888739797804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,1024,0.004089777668317159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,1536,0.014020444618331062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,768,0.0037271111375755737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,1024,0.013822221921549903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,512,0.003412444351447953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,768,0.01349688900841607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,256,0.003060444361633725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,512,0.013381332986884646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,128,0.0028844444702068963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,256,0.01295644458797243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,64,0.0027395555128653846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,128,0.012862222062216865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,32,0.002768888862596618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,64,0.012657778130637275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,32,0.012451555165979596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,1024,0.029398222764333088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,768,0.028237332900365193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,512,0.027663111686706543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,256,0.026038222842746313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,128,0.025594666600227356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,16384,0.38382932874891496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,10240,0.24720001220703125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,12288,0.2929813332027859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,16384,0.4263751241895888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,12288,0.3585804568396674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,10240,0.47705687416924375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,12288,0.5707786877950033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,10240,0.3021431234147814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,8192,0.2016835610071818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,8192,0.3832017845577664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,7168,0.19963467121124268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,6144,0.1567022270626492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,7168,0.22158043914371064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,8192,0.24860265519883898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,16384,0.7500489022996691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,6144,0.28846311569213867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,6144,0.19487644566429985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,7168,0.3363635540008545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,5120,0.13422932889726427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,5120,0.16899377769894072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,4096,0.12303645080990261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,5120,0.2409528891245524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,3584,0.10095822148852879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,4096,0.14288711547851562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,4096,0.19531822204589844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,3584,0.12710311677720812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,3072,0.08994488583670722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,3072,0.10477688577440049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,3584,0.20039110713534883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,3072,0.14857510725657144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,2560,0.07714755667580499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,2048,0.06551822026570638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,1536,0.05868266688452827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,2560,0.10390400224261814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,2048,0.0898231135474311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,1536,0.0709644423590766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,1536,0.07761955261230469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,1024,0.04478044311205546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,2560,0.12540354993608263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,768,0.040207998620139226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,1024,0.05827199750476413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,768,0.05163288778728909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,768,0.042760888735453285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,2048,0.10025155544281006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,512,0.03050133254792955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,512,0.03706489006678263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,256,0.031593776411480375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,128,0.010299555957317352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,1024,0.053336888551712036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,128,0.02915999955601162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,256,0.03688888748486837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,512,0.045351998673544995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,64,0.007479999628331926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,256,0.017208000024159748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,128,0.033450666401121355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,64,0.026742221580611333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,32,0.005892444401979446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,32,0.027639999985694885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,12288,0.08884177605311076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,16384,0.11230399873521592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,16384,0.13684711191389295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,12288,0.1405280033747355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,16384,0.18547111087375215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,12288,0.11208000448015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,10240,0.0817413330078125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,8192,0.06484711170196533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,10240,0.11744622389475505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,10240,0.10233689016766018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,8192,0.09546044137742783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,8192,0.081185777982076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,7168,0.0588595536020067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,7168,0.07582133346133761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,6144,0.05327199896176656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,7168,0.08404266834259033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,6144,0.07331466674804688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,6144,0.06922399997711182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,65536,0.4006275600857205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,5120,0.04765422145525614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,5120,0.06484888659583198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,5120,0.061928888161977134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,3584,0.03855999973085191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,4096,0.04138044516245524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,4096,0.06151022513707479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,4096,0.0503386656443278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,3584,0.04483022292455038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,3072,0.03566755519972907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,3072,0.05119199885262383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,3584,0.055063111914528735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,3072,0.03901422354910109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,2560,0.037795556916130915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,2560,0.03296622302797105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,2048,0.029318223396937054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,1536,0.02587466604179806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,2560,0.049343112442228526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,65536,0.4512951109144423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,2048,0.027242667145199243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,2048,0.04563111066818237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,1536,0.04134933484925164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,1536,0.021565333008766174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,768,0.01203555530971951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,1024,0.02360533343421088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,768,0.020947555700937908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,512,0.009104000197516548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,256,0.006261333409282897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,512,0.01962311069170634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,768,0.035255110926098294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,256,0.015498666299713982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,1024,0.015271999769740634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,128,0.003942222230964237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,512,0.03295466634962294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,64,0.003656889001528422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,256,0.027386666999922857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,32,0.00387999994887246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,128,0.015563555889659457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,128,0.026542221506436665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,1024,0.03702844513787164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,32,0.017588444881969027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,64,0.016252444850073922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,16384,0.09139821926752727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,16384,0.1094320019086202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,12288,0.11311466164059109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,16384,0.1506399975882636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,12288,0.07385422123803033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,10240,0.06362222300635444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,12288,0.09098311265309651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,10240,0.0944826669163174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,10240,0.08113688892788358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,8192,0.0539244446489546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,8192,0.07666044765048556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,8192,0.07761955261230469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,7168,0.06765155659781562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,7168,0.049477332168155246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,7168,0.06647021902932061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,6144,0.044601778189341225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,6144,0.05836177534527249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,6144,0.06191911300023397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,5120,0.04064711266093784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,5120,0.04899911085764567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,65536,0.319106658299764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,5120,0.057612445619371205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,4096,0.03579644362131754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,4096,0.05260622170236376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,4096,0.04079822368092007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,3584,0.03344800074895223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,3584,0.036093334356943764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,3584,0.0498835543791453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,3072,0.03145422206984626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,2560,0.0303457776705424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,3072,0.031304889255099826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,3072,0.046893331739637584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,2560,0.026891556051042344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,2560,0.04528799984190199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,65536,0.3948204517364502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,2048,0.025623111261261836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,2048,0.04154488775465224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,1024,0.012607110871209038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,1536,0.024691555235120986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,1536,0.038375112745496966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,1024,0.021126222279336717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,768,0.010025777750545079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,768,0.020163555939992268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,1536,0.017484444710943434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,1024,0.0351857774787479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,512,0.00775911079512702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,2048,0.02239111065864563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,512,0.017670222454600863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,256,0.005374222166008419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,768,0.034056001239352755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,512,0.032002667586008705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,256,0.015232000086042615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,128,0.0038417776425679526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,65536,0.749175124698215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,128,0.015285332997639975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,256,0.02758666707409753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,64,0.0034186666210492453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,128,0.026464000344276428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,64,0.01572088897228241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,32,0.0035208890007601846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,32,0.016383111476898193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,16384,0.0825528899828593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,16384,0.09857955243852402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,65536,0.5762711101108128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,12288,0.08977155552970038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,16384,0.11765778064727783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,12288,0.06621155473921034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,12288,0.08252533276875813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,10240,0.05808089176813761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,10240,0.07480266359117296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,10240,0.07541866434945001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,8192,0.04980533321698507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,8192,0.061371558242373996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,8192,0.06517955329683092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,7168,0.04972088999218411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,7168,0.05387733379999796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,6144,0.04183200001716614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,7168,0.06181866592831082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,6144,0.05699466996722751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,5120,0.03996710975964864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,6144,0.04745422138108147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,5120,0.03779200050565932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,4096,0.03302399979697333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,5120,0.053321778774261475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,4096,0.0329422222243415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,3584,0.030984888474146526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,4096,0.049040890402264066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,3072,0.028704888290829126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,3584,0.04671733246909248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,65536,0.28210221396552193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,65536,0.33627822664048934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,3584,0.029226667351192896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,3072,0.04419733418358696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,3072,0.025877333349651758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,2560,0.026410667432679072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,2048,0.02440266642305586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,1536,0.014476444986131458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,2560,0.042824887567096286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,1536,0.022643556197484333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,2048,0.03966222206751505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,1024,0.010604444477293225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,1536,0.03729155659675598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,1024,0.020780444145202637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,768,0.00887022250228458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,2560,0.022120889690187242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,2048,0.018343110879262287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,768,0.01962311069170634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,512,0.007167110840479533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,768,0.033591111501057945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,512,0.01632000009218852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,512,0.031098667118284438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,256,0.015070221490330167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,1024,0.03549777799182468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,256,0.004040889028045866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,128,0.003689777933888965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,256,0.027169777287377253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,128,0.015077332655588785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,64,0.003408000080121888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,128,0.026362667481104534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,32,0.0034053333931499054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,64,0.01534844438234965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,32,0.016001777516471017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,16384,0.06763466861512926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,16384,0.08619022369384766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,16384,0.0933235552575853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,65536,0.48296176062689883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,12288,0.05523288912243313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,12288,0.07183911403020223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,12288,0.07347466548283894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,10240,0.06018844578001234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,10240,0.04822311136457655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,8192,0.04173688756095039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,10240,0.06670400169160631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,8192,0.04903466502825419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,8192,0.05945510996712578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,7168,0.03839377893341912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,7168,0.04353422257635328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,6144,0.035233779086007014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,7168,0.0559555557039049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,6144,0.038351111941867404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,6144,0.052336888180838696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,5120,0.03194400005870395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,65536,0.22210578123728433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,5120,0.03251377741495768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,5120,0.04946488804287381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,65536,0.2841004530588786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,3584,0.025569778349664476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,4096,0.027702222267786663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,3584,0.04343466626273262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,3072,0.023621333969963923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,4096,0.045528001255459256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,4096,0.026838221483760413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,2560,0.0222062220176061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,3072,0.04001422060860528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,2560,0.04040266738997565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,2048,0.020788444413079154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,3072,0.021086222595638696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,3584,0.02387822171052297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,1536,0.011908444265524546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,2048,0.03767555620935228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,1536,0.01941866676012675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,2048,0.016711110870043438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,1024,0.008960888617568547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,2560,0.018052443861961365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,768,0.007635555333561367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,1024,0.01772888832622104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,1536,0.037046220567491316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,768,0.015214223000738354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,512,0.006171555568774541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,1024,0.03388444582621256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,768,0.032687111033333674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,128,0.0035439998739295532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,512,0.014114666316244336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,256,0.00387999994887246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,512,0.028208888239330713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,256,0.013528889252079858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,256,0.02627022233274248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,128,0.013639999760521783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,64,0.003419555516706573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,128,0.026130666335423786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,32,0.0033537778589460584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,64,0.013738666971524557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,32,0.014462222655614218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,16384,0.06337244643105401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,16384,0.08164711130989923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,12288,0.06359644730885823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,12288,0.051312890317704946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,12288,0.07102311319775052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,16384,0.08888355228636001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,10240,0.044906665881474815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,10240,0.053447110785378345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,8192,0.03953155544069078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,10240,0.06373688909742567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,8192,0.05583555830849541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,8192,0.04351910948753357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,7168,0.03602755400869582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,7168,0.038626667526033186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,6144,0.033152000771628484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,7168,0.05375733309321933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,65536,0.3878657817840576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,6144,0.050556444459491305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,6144,0.03418044580353631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,5120,0.030142221185896132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,65536,0.20948710706498888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,65536,0.26154666476779515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,4096,0.02643199927277035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,5120,0.028936889436509874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,4096,0.043776889642079674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,5120,0.047819554805755615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,3584,0.02402666707833608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,4096,0.024136889312002394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,3584,0.041645334826575384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,3072,0.022429333792792425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,2560,0.016374222106403775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,2560,0.021365332934591506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,3072,0.04103111227353414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,2048,0.013627555635240344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,2048,0.019983111156357657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,2560,0.03919822308752272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,3072,0.019116444720162284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,2048,0.03729866610633002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,1536,0.010951110886202918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,1536,0.018564444449212816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,1024,0.008401778009202745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,1536,0.035522666242387556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,1024,0.017179555363125272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,3584,0.02425777746571435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,768,0.007165333463086023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,1024,0.033238222201665245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,768,0.01571466690964169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,512,0.005344888816277186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,768,0.03129244512981839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,512,0.014090667168299357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,256,0.0036675557494163513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,512,0.027980443504121568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,256,0.013495999905798169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,128,0.0034106667670938703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,256,0.026268444127506677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,128,0.013589333328935834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,64,0.003245333416594399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,128,0.026174222429593403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,32,0.0033679999825027254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,64,0.013618666264745923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,32,0.014383110735151501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,16384,0.05824089050292969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,16384,0.07577421930101183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,16384,0.08226666847864787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,12288,0.047135111358430654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,12288,0.06514488988452487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,12288,0.057982219590081104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,10240,0.04194577866130405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,10240,0.04874666531880697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,10240,0.059827552901373975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,8192,0.05298577745755514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,7168,0.033591111501057945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,8192,0.03656533360481262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,65536,0.1862835619184706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,65536,0.3515360090467665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,7168,0.05108888943990072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,8192,0.040688001447253756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,7168,0.03545422355333964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,5120,0.027802666028340656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,6144,0.030902223454581365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,5120,0.0455662210782369
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,4096,0.024243555135197107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,5120,0.02628799941804674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,4096,0.04161244299676683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,65536,0.23062311278449166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,6144,0.050316443045934044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,6144,0.031187554200490315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,3584,0.023206222388479445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,3072,0.021479111578729417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,3584,0.03993955585691664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,2560,0.015086222026083203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,4096,0.022321777211295232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,3072,0.03782044516669379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,3584,0.01993333299954732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,2560,0.020639111598332722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,2048,0.012725333372751871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,2560,0.03792622354295518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,3072,0.017454221844673157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,2048,0.019328888919618394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,1536,0.010436444646782346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,1536,0.01831022236082289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,2048,0.03641688823699951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,1024,0.007885333564546373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,1024,0.016623111234770883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,1536,0.0351875556839837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,768,0.006739555547634761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,768,0.014688000082969666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,1024,0.033225778076383806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,512,0.004716444346639845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,768,0.029370665550231934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,512,0.013984888792037964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,256,0.003658666792843077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,512,0.02787999974356757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,256,0.013521778086821238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,128,0.003359111025929451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,256,0.026884444885783728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,128,0.013341333303186627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,64,0.003198222153716617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,128,0.026152888933817547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,32,0.0032675556010670135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,64,0.013528000149461957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,32,0.014279999666743808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,16384,0.07006755802366468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,16384,0.05126844512091743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,12288,0.04670844475428263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,16384,0.06642399893866645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,12288,0.042544888125525586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,12288,0.061362663904825844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,10240,0.06019022067387899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,10240,0.03791288865937127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,8192,0.0333315564526452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,10240,0.03963377740648057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,8192,0.05048177639643351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,7168,0.03084889054298401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,7168,0.02908088763554891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,8192,0.03237777617242601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,6144,0.028104000621371802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,7168,0.04898400108019511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,65536,0.2884604401058621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,6144,0.04570400052600437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,5120,0.02199288871553209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,5120,0.025595555702845257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,6144,0.025663110944959853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,4096,0.01824977828396691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,4096,0.022802667485343084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,5120,0.042965332667032875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,4096,0.04059999850061204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,3584,0.016417778200573392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,3584,0.021540444758203294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,3072,0.014431110686726041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,3584,0.03903111153178745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,3072,0.020530665914217632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,2560,0.012477333347002665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,65536,0.1581759982638889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,3072,0.037235554721620344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,2048,0.010616000327799054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,2560,0.01975644462638431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,2048,0.018636445204416912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,2560,0.03717066513167487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,1536,0.008930666579140557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,65536,0.1990062263276842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,2048,0.035967999034457736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,1536,0.017834666702482436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,1024,0.0069439998931354946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,1024,0.014742222097184924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,1536,0.03492710987726847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,768,0.006238222122192383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,768,0.014521777629852295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,1024,0.031194667021433514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,512,0.013909333282046847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,512,0.003964444415436851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,768,0.02924355533387926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,256,0.003585777762863371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,512,0.02793866727087233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,128,0.0033084443873829315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,256,0.013391999734772576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,256,0.026191999514897663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,128,0.02593155536386702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,128,0.013362666798962487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,64,0.00313688897424274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,32,0.003181333343187968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,64,0.013407111167907715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,32,0.014000000225173103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,16384,0.04580266608132256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,16384,0.06421955426534016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,12288,0.03842488924662272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,16384,0.05283733208974203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,12288,0.04191022117932638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,10240,0.034183998902638756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,12288,0.05629511011971367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,10240,0.03580800029966567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,10240,0.05248622099558512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,8192,0.03053155541419983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,65536,0.1382711066140069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,7168,0.02786933382352193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,8192,0.029896001021067303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,8192,0.04653777678807577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,7168,0.04529955652025011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,6144,0.02557955516709222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,6144,0.04182044333881802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,7168,0.027253333065244887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,5120,0.023194667365815904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,5120,0.018561777141359117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,4096,0.02092444399992625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,5120,0.04021866785155402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,65536,0.17516088485717773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,4096,0.03837777839766608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,65536,0.23570489883422852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,3584,0.013519110778967539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,3584,0.020305777589480083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,6144,0.021513778302404616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,3072,0.012267555627557965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,4096,0.015223999818166098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,2560,0.010630221830474006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,3584,0.03734755516052246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,3072,0.020404444800482858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,2560,0.018815999229749043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,2048,0.009232000344329411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,3072,0.03554755449295044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,2048,0.0177848885456721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,2560,0.03606755534807841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,1024,0.00620977787507905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,1536,0.00795377790927887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,2048,0.035418666071361966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,1536,0.016452444924248587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,1536,0.034281777011023626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,768,0.005528000079923206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,1024,0.014567111929257711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,768,0.029241777128643457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,768,0.014295111099878946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,1024,0.029603554142846003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,512,0.003809777812825309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,512,0.013903111219406128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,256,0.003407111184464561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,256,0.013372444444232516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,128,0.003162666741344664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,512,0.027619555592536926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,128,0.013268444273206921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,256,0.026735999517970618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,64,0.003007111036115222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,128,0.026191110412279766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,32,0.003170666595300039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,64,0.01332622187005149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,32,0.013825777504179211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,16384,0.04301777813169691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,16384,0.06162133481767443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,16384,0.04750755429267883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,12288,0.036454220612843834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,12288,0.05460266603363884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,12288,0.037704888317320086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,10240,0.032716443141301475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,8192,0.028390222125583228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,65536,0.19067377514309355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,10240,0.050599998897976346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,8192,0.045260445939169995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,10240,0.03218577636612786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,7168,0.025938666529125635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,65536,0.13025689125061035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,8192,0.027241778042581346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,7168,0.04376888937420315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,6144,0.024010666542583044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,5120,0.016504888733228047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,5120,0.022117333279715642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,6144,0.04056533177693685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,5120,0.040114667680528425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,4096,0.013605333036846586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,7168,0.024744888146718342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,4096,0.020632889535692003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,3584,0.012230222423871359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,6144,0.019327110714382596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,4096,0.03775377737151252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,3072,0.011215111447705163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,2560,0.009719111025333405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,3584,0.020025778147909377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,2560,0.018607111440764535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,3072,0.01920000049802992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,3072,0.03513688842455546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,3584,0.03672444489267137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,2048,0.008577778107590145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,2560,0.03579644362131754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,2048,0.017463111215167575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,1536,0.00720977783203125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,2048,0.03520355621973673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,1536,0.015406222807036506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,1024,0.014547554983033074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,1536,0.03378933336999681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,1024,0.006106666806671355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,768,0.0046471113132105935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,1024,0.02936977810329861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,512,0.0037439999481042228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,768,0.014341332846217685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,65536,0.16375022464328343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,512,0.02774577836195628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,768,0.02903911140229967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,512,0.013730666703648038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,256,0.0033644443998734155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,256,0.013369777964221107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,128,0.003157333367400699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,256,0.026131555438041687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,128,0.01310755560795466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,64,0.002924444360865487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,128,0.02629599968592326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,32,0.0030168888883458245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,64,0.013264888690577613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,32,0.0138586668504609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,16384,0.04148355457517836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,12288,0.03494577937655979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,12288,0.05201955636342367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,16384,0.058119111590915255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,10240,0.030883553955290053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,12288,0.03373333480623033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,10240,0.047990222771962486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,16384,0.043176889419555664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,10240,0.02934755550490485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,8192,0.02702933384312524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,7168,0.02289955483542548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,8192,0.04376177655325996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,65536,0.17557777298821342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,7168,0.024811555941899616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,8192,0.024831111232439678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,7168,0.04220088985231188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,6144,0.02353511088424259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,6144,0.03966133462058173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,5120,0.017474666237831116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,5120,0.021776000658671062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,65536,0.12558221817016602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,4096,0.012277333272827996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,5120,0.03927910990185208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,3584,0.011287111375066968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,4096,0.02015822297996945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,6144,0.020433778564135235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,4096,0.037421332465277776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,3584,0.01942488882276747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,3072,0.010258666343159145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,3584,0.03647289011213515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,3072,0.01903199983967675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,3072,0.03525244527392917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,2560,0.009053332938088311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,2560,0.017990221579869587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,2048,0.007866666548781926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,2560,0.03593688872125413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,1536,0.007044444481531779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,2048,0.01633777717749278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,2048,0.03487644592920939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,1536,0.01587466730011834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,65536,0.15676355361938477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,1024,0.005527110977305307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,1536,0.03252089023590088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,1024,0.014656000667148165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,768,0.014200000299347771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,768,0.004351110921965705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,1024,0.029594666428036157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,512,0.0037235555549462638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,768,0.02887644370396932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,256,0.0033760000434186724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,512,0.013819555441538492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,256,0.013463999662134381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,128,0.003148444410827425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,512,0.027585777971479628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,256,0.0262435558769438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,128,0.01313066648112403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,64,0.003010666618744532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,128,0.026016000244352553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,32,0.0030382221771611106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,64,0.013225778109497495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,32,0.013617777989970313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,16384,0.039868444204330444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,16384,0.056504891978369824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,12288,0.03345688846376207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,16384,0.03570844398604499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,12288,0.02872177627351549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,10240,0.02995377779006958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,12288,0.049991998407575816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,10240,0.046422223250071205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,10240,0.02500444485081567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,8192,0.02564711040920681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,8192,0.021107556091414556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,8192,0.04135466615358988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,65536,0.1453919940524631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,7168,0.0244159996509552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,6144,0.01717333330048455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,7168,0.040771553913752236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,6144,0.022827555735905964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,5120,0.015013333823945789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,6144,0.038690666357676186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,7168,0.01925066610177358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,5120,0.021528888079855178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,4096,0.013013333082199097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,4096,0.020066666934225295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,5120,0.038303110334608294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,3584,0.012372444073359171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,3584,0.019132445255915325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,4096,0.03685955537690057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,65536,0.1126248836517334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,3072,0.009242666264375051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,3584,0.035976890060636736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,3072,0.018632888793945312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,2560,0.008526222573386298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,2560,0.017474666237831116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,3072,0.03444622291458978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,2048,0.007295110987292395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,2560,0.035150223308139376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,2048,0.015808888607554965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,1536,0.006623999940024481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,2048,0.03259822063975864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,1536,0.015035554766654968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,1024,0.0053173332578606075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,65536,0.14437688721550837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,1024,0.029463112354278564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,1536,0.031162665949927434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,1024,0.014475555883513557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,768,0.028958221276601154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,768,0.004477333277463913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,768,0.014127110441525778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,512,0.0037048889531029593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,512,0.013838222457302941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,256,0.0033537778589460584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,512,0.02769777841038174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,128,0.013297778036859302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,128,0.00309955556359556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,256,0.013358222113715278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,256,0.025989333788553875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,64,0.0029351111087534162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,128,0.026221333278550044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,32,0.00301155551440186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,64,0.013191111385822296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,32,0.013536888692114087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,16384,0.05432266659206814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,12288,0.03173155585924784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,16384,0.0381040010187361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,12288,0.023935111032591924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,12288,0.048872000641293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,65536,0.12455644872453477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,10240,0.028419554233551025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,16384,0.02940800123744541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,10240,0.046478221813837685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,8192,0.02492888934082455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,7168,0.01608800060219235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,8192,0.041357331805759005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,10240,0.02126933303144243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,7168,0.023504000571038988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,6144,0.01462577780087789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,8192,0.01758577757411533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,7168,0.04067022270626492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,6144,0.022143999735514324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,65536,0.11411110560099284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,5120,0.013219555219014486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,6144,0.03838488790724013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,5120,0.020996444755130347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,4096,0.011552888486120435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,3584,0.010888888604111142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,5120,0.03796266516049703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,4096,0.01966399947802226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,3584,0.01887999971707662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,4096,0.0367253323396047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,65536,0.14519466294182673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,3072,0.008772444393899705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,3584,0.03571733170085483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,3072,0.01722666621208191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,2560,0.008068444828192392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,3072,0.03423377871513367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,2560,0.016851555969980028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,2048,0.006958222223652734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,2560,0.033624887466430664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,2048,0.015627556376987033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,1536,0.006246222390068903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,2048,0.032448000378078885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,1536,0.015037332971890768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,1024,0.004620444443490771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,1536,0.031222220924165513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,1024,0.014350222216712104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,768,0.0040702223777771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,1024,0.028929776615566675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,768,0.014183110660976835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,512,0.0036720000207424164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,768,0.028558221128251817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,512,0.013623110949993134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,256,0.00328177772462368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,256,0.013260444833172692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,512,0.027616888284683228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,128,0.003039999968475766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,256,0.025957332717047796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,128,0.025811556312772963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,128,0.012934221989578672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,64,0.002903999967707528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,32,0.0029404444826973807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,64,0.013218666944238873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,32,0.01352355546421475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,16384,0.03594133257865906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,12288,0.019483556350072224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,16384,0.05170933405558268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,65536,0.10211555825339423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,12288,0.030132446024152968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,12288,0.046711110406451754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,10240,0.016936888297398884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,10240,0.0269795556863149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,8192,0.014683556225564746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,10240,0.0433991120921241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,8192,0.024438222249348957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,8192,0.03971733318434821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,16384,0.024272888898849487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,7168,0.01365600029627482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,7168,0.0229066660006841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,6144,0.012528888881206512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,6144,0.020890666378868952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,7168,0.03940533267127143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,5120,0.011590222517649332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,6144,0.037943111525641546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,5120,0.02050133380624983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,4096,0.009340444372759925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,5120,0.03746933407253689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,4096,0.0188080006175571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,3584,0.008811555802822113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,3584,0.017165333032608032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,3072,0.008026666939258575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,4096,0.03611288799179925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,3584,0.03459199931886461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,3072,0.016747555798954435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,2560,0.007418666448858049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,3072,0.033187554942237005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,2048,0.0063164445261160536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,65536,0.13248444928063288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,2560,0.015871110889646742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,1536,0.005103999955786599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,65536,0.10782933235168457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,2048,0.015201777219772339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,2560,0.032933334509531655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,1536,0.030927111705144245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,2048,0.03141599893569946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,1024,0.004362666772471534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,1536,0.014783110883500842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,1024,0.014118222726715935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,768,0.004032888760169347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,768,0.014087999860445658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,1024,0.029021332661310833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,512,0.003663111064169142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,512,0.027494221925735474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,768,0.02882222334543864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,512,0.013656888571050433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,256,0.0032924444725116095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,128,0.01293244461218516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,256,0.01331377774477005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,128,0.0030817778574095834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,256,0.02585244509908888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,65536,0.0765226682027181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,64,0.0029680000411139596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,128,0.025961778230137292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,32,0.0029564443975687027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,64,0.01293955577744378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,32,0.013455111119482251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,16384,0.02784622377819485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,12288,0.014752000570297241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,16384,0.05096711052788628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,12288,0.024384000235133704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,10240,0.013463999662134381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,12288,0.045900444189707436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,10240,0.02236533330546485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,16384,0.01850755512714386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,8192,0.011709333293967776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,10240,0.042839109897613525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,8192,0.02090488870938619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,7168,0.011251555548773872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,8192,0.0396728879875607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,7168,0.01998044384850396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,6144,0.010088889135254754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,65536,0.07628266678916083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,7168,0.039064000050226845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,6144,0.019125334090656705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,5120,0.0103493332862854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,6144,0.03706666827201843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,5120,0.01720266706413693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,4096,0.009476444787449306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,4096,0.01646133263905843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,5120,0.036398222049077354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,3584,0.00909511082702213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,3584,0.015992888145976596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,4096,0.033703111939960055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,3072,0.008432000047630733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,65536,0.12484622001647949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,2560,0.0069155556460221606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,3072,0.015598222613334656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,3584,0.032835556401146784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,2560,0.015421332584487068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,2048,0.006132444573773279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,3072,0.03145600027508206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,2560,0.03216177887386746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,2048,0.01482311056719886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,1536,0.005015999906592899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,65536,0.05624000231424967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,2048,0.0310808883772956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,1536,0.014398222168286642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,1024,0.004272888931963178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,1536,0.03051111102104187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,1024,0.014031110538376702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,768,0.003957333250178231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,1024,0.02885422110557556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,768,0.014017777310477363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,512,0.0036355555057525635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,768,0.02851466668976678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,512,0.0135057775510682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,256,0.0032755554550223877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,512,0.02713955607679155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,128,0.003064888839920362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,256,0.013184888495339287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,256,0.025814221964942083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,128,0.013251555462678274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,64,0.0029022221763928733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,128,0.025593777497609455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,32,0.0029493332323100832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,64,0.012806221842765808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,32,0.013200889031092325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,16384,0.014503111441930136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,16384,0.026682666606373254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,12288,0.012845333251688214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,12288,0.02383022175894843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,16384,0.050085332658555776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,10240,0.011997333003415002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,12288,0.045824001232783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,10240,0.021942221456103857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,8192,0.011696889168686338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,10240,0.042838222450680204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,8192,0.019967110620604623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,7168,0.011057777537239922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,7168,0.01923733287387424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,65536,0.07315999931759305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,6144,0.010430222584141625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,8192,0.03959822323587205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,6144,0.01767200065983666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,7168,0.03892622060245938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,5120,0.009791110952695211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,6144,0.03626399901178148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,5120,0.016880000631014507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,4096,0.00847822262181176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,65536,0.1242648892932468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,5120,0.03695911169052124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,65536,0.04647377795643277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,4096,0.016597333881590102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,3584,0.00831733312871721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,4096,0.03340977761480544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,3584,0.015927111109097798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,3072,0.03164711263444688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,3072,0.007558222446176741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,3584,0.03221244282192654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,3072,0.015580443872345818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,2560,0.006707555717892117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,2560,0.015322667029168872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,2048,0.005741333381997214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,2048,0.014920888675583733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,2560,0.031694223483403526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,1536,0.004895110925038655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,2048,0.0306693348619673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,1536,0.014533332652515836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,1024,0.004201777693298128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,1024,0.028976887464523315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,1536,0.03019733230272929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,1024,0.014152000347773233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,768,0.028312000963422988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,768,0.0038906666967603895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,768,0.013768889009952545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,512,0.027144889036814373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,512,0.003532444437344869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,128,0.0030053332448005676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,512,0.013434666726324292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,256,0.003257777748836411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,256,0.013174222575293647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,256,0.026178666286998328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,64,0.0028488888508743713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,128,0.01295555548535453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,128,0.025776889589097764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,32,0.0028799999919202593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,64,0.01277599980433782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,32,0.013163555827405719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,16384,0.0118231111102634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,16384,0.026567111412684124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,12288,0.011909333368142446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,16384,0.04958933260705736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,12288,0.023561777340041265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,10240,0.01108088923825158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,10240,0.021320889393488567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,8192,0.009993777506881291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,12288,0.046015110280778676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,65536,0.06896177927652995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,10240,0.04277688927120633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,8192,0.01886399918132358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,7168,0.009499555660618676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,6144,0.008829333715968663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,65536,0.12486488289303249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,7168,0.01811733345190684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,8192,0.03927288783921136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,7168,0.037406222687827215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,6144,0.017242666747834947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,5120,0.007856888903511895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,6144,0.0347724457581838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,5120,0.016549333930015564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,4096,0.007239110767841339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,65536,0.03438933359252082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,5120,0.035262223747041486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,4096,0.01604711181587643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,3584,0.007654222349325816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,4096,0.03328355484538608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,3584,0.015437333120240105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,3072,0.007058666812049017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,3584,0.032144000132878624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,3072,0.015245333313941956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,2560,0.0061919999619325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,3072,0.031170666217803955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,2048,0.005605333381228977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,2048,0.031034668286641438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,2560,0.014740443891949125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,2048,0.014469333820872836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,2560,0.031559109687805176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,1536,0.004860444615284602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,1536,0.014112000664075216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,1024,0.004159111115667555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,1024,0.01384800010257297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,1536,0.030452443493737116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,768,0.0038213332494099936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,1024,0.028573334217071533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,768,0.013741333451535968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,512,0.0035137778355015647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,768,0.028288000159793433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,512,0.013426666458447775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,256,0.0032142222755485107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,512,0.027311111489931744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,256,0.013069333301650153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,64,0.013036444783210754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,128,0.002999111182159848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,256,0.025736888249715168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,128,0.012995555169052549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,64,0.0028275555620590844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,128,0.02555466691652934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,32,0.0028355556229750314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,32,0.013146667016877068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,16384,0.010326222413116032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,65536,0.06744266880883111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,16384,0.025271111064487036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,12288,0.00885333369175593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,65536,0.1251253286997477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,16384,0.049264887968699135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,12288,0.021711111068725586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,10240,0.008633777499198914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,12288,0.04552177919281853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,65536,0.02269866731431749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,10240,0.01997955474588606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,8192,0.007558222446176741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,10240,0.040919999281565346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,8192,0.018735110759735107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,7168,0.007199111084143321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,7168,0.017997332745128207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,8192,0.03672000103526645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,6144,0.006824000014199152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,7168,0.03649155630005731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,6144,0.017204445269372728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,5120,0.00721777809990777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,6144,0.034747554196251765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,5120,0.016604445046848722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,4096,0.00683822234471639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,5120,0.03415822320514255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,4096,0.015849777393870883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,3584,0.007313778003056844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,4096,0.03295733200179206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,3584,0.015396444333924187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,3072,0.006991110742092133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,2560,0.006174222048785951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,3584,0.03229866756333245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,3072,0.015125332607163323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,2560,0.031770666440327965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,3072,0.030826667944590252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,2560,0.014870221416155497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,2048,0.005483555710977978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,2048,0.014426666829321118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,1536,0.004867555366622077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,2048,0.030878222650951807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,1536,0.01421955558988783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,1024,0.004126222183307012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,1536,0.030023111237419978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,1024,0.013726222018400827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,768,0.0038328886859946777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,1024,0.028611554039849177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,768,0.013819555441538492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,512,0.0035013332962989807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,512,0.02697511182890998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,768,0.028151111470328435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,512,0.013329777452680798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,256,0.0031662223239739737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,256,0.013042666845851474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,128,0.002954666606254048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,128,0.01293155550956726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,256,0.026473777161704168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,64,0.0028115556471877625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,128,0.025892444782786902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,32,0.002830222249031067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,64,0.012689777546458773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,32,0.012925333446926542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,65536,0.014426666829321118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,65536,0.06553600231806438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,16384,0.00737955586777793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,12288,0.007003555695215861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,16384,0.024511110451486375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,65536,0.12459288703070746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,12288,0.021549334128697712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,10240,0.007253333098358578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,16384,0.047953777843051486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,12288,0.04359199934535556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,10240,0.019928889142142404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,8192,0.03644888930850559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,8192,0.007040888898902469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,10240,0.04018311036957635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,8192,0.018592000007629395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,5120,0.007160888777838812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,7168,0.0069848886794514125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,7168,0.01791466606987847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,6144,0.006780444333950679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,7168,0.03621066610018412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,6144,0.01711466742886437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,6144,0.0344524449772305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,5120,0.016417778200573392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,4096,0.006768888897365994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,5120,0.03401244348949856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,4096,0.015855110353893705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,3584,0.00720266666677263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,4096,0.03330399923854404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,3584,0.015368888775507608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,3072,0.0068986668354935115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,2560,0.01475644442770216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,3584,0.03200444579124451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,3072,0.015199111567603217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,2560,0.006165333506133821
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,3072,0.030733333693610296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,2048,0.005450666778617435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,2560,0.03167377909024557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,1536,0.0048071112897661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,2048,0.014466666513019137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,2048,0.030631999174753826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,1536,0.01426755554146237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,1024,0.004173333446184794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,512,0.003576888806290097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,1536,0.03012266755104065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,1024,0.013807999591032663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,512,0.013425778183672162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,768,0.0038293335172865125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,1024,0.028619554307725694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,768,0.013624000052611033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,768,0.027930667002995808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,256,0.0031271111220121384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,512,0.02735911144150628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,256,0.01312800000111262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,128,0.0029440000653266907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,256,0.025935999221271936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,128,0.012791111237472959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,64,0.0028071111689011254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,32,0.0028035555862718155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,128,0.025507556067572698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,64,0.013010666602187686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,32,0.013025778035322825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,65536,0.012994666894276937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,16384,0.007187555233637492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,65536,0.0657786660724216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,12288,0.006804444309737947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,16384,0.024551999237802293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,10240,0.007145778172545963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,12288,0.021573333276642695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,8192,0.006919999917348226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,10240,0.019884443945354886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,7168,0.006828444285525217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,8192,0.01849600010448032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,6144,0.006720888945791457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,7168,0.017872888180944655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,5120,0.007163555257850223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,6144,0.01715377800994449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,4096,0.00674933319290479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,5120,0.016294222739007738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,3584,0.007096000015735626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,4096,0.01570933394961887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,3072,0.006868444383144379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,2048,0.0054551110499435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,3584,0.01533333294921451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,1536,0.004826666580306159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,2560,0.006114666660626729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,3072,0.01516266663869222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,2560,0.014679999815093147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,1024,0.013773333695199756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,2048,0.0143857780430052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,768,0.013821333646774292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,1024,0.004143110993835661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,512,0.013246222502655454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,256,0.003118222165438864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,1536,0.014108444253603617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,768,0.0037573331760035623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,512,0.0034515555534097883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,128,0.002941333378354708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,256,0.013270222478442721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,64,0.0028115556471877625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,128,0.012830221818553077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,32,0.0028071111689011254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,64,0.012659555508030785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,32,0.01293244461218516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,65536,0.011880888707107969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,16384,0.006888000087605582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,12288,0.006791111081838608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,65536,0.06640711095598009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,16384,0.024442666106753882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,10240,0.00703999979628457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,12288,0.02141333288616604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,8192,0.00684533350997501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,10240,0.01998844411638048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,7168,0.006784889019197888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,8192,0.018527110417683918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,6144,0.006673777682913675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,7168,0.01795022189617157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,5120,0.0070568886068132185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,6144,0.017272000511487324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,4096,0.00665244460105896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,5120,0.016543111867374845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,3584,0.007095110913117726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,4096,0.015640000502268474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,3584,0.015297777122921415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,3072,0.006792000184456508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,2560,0.006102222121424145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,3072,0.014816888504558139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,2048,0.005415110952324337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,2560,0.014650666051440768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,1536,0.0047351109484831495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,2048,0.014358222484588623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,1024,0.0041422223051389056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,1536,0.01406222250726488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,768,0.003710222327046924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,1024,0.013725333743625216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,512,0.0034213333080212274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,768,0.013577777478430005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,256,0.003084444544381566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,512,0.013279111021094851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,128,0.0028604444944196274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,256,0.013255111045307584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,64,0.0027724444452259275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,128,0.012758221891191272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,32,0.0027822222974565295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,64,0.012598222328556908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,32,0.012879999975363413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,12288,0.2952248785230848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,10240,0.25157154930962455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,12288,0.35314133432176376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,16384,0.3825751145680745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,8192,0.20737511581844756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,16384,0.42678578694661456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,10240,0.4757937855190701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,10240,0.2978755633036296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,8192,0.24719111124674478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,8192,0.3819982210795085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,12288,0.567671087053087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,6144,0.16010222170088026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,7168,0.18683200412326387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,7168,0.22105777263641357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,7168,0.3323813279469808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,5120,0.13418043984307185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,6144,0.19444621933831108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,6144,0.28718045022752553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,16384,0.7502791086832682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,4096,0.11017333136664496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,3584,0.10961600144704182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,4096,0.1363146702448527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,5120,0.16745600435468885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,5120,0.23997333314683703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,3584,0.19918489456176758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,4096,0.19419556193881563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,3072,0.08899377452002631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,3584,0.12892799907260472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,2560,0.08396177821689182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,3072,0.11554755104912652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,3072,0.14658489492204455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,2048,0.06540622313817342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,1536,0.05374222331576877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,2560,0.09495111306508382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,2048,0.08108177449968126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,1536,0.0692159997092353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,2560,0.1236071056789822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,1024,0.04402577877044678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,1024,0.057322667704688184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,1024,0.05236088898446825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,768,0.039860443936453924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,256,0.01645511057641771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,2048,0.09888533088896011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,256,0.03031733300950792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,1536,0.07637600104014079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,128,0.009663110805882348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,512,0.03587111168437534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,256,0.037273777855767146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,512,0.02918400035964118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,128,0.028247111373477515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,64,0.00703288863102595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,32,0.005375999957323074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,512,0.04400622182422214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,768,0.05036355720625984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,768,0.04142399960094028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,64,0.026157332791222468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,128,0.03311289019054837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,32,0.026084444589085046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,12288,0.0886933339966668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,16384,0.11208621660868327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,16384,0.14211911625332302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,12288,0.14166577657063803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,12288,0.11148088508182102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,10240,0.07667111025916205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,10240,0.09244000249438816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,16384,0.18818399641248915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,10240,0.11810755729675293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,8192,0.06436000267664592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,8192,0.07911466889911227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,7168,0.060552000999450684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,8192,0.09584710995356242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,7168,0.0748515526453654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,7168,0.08442222409778172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,6144,0.07382666402392916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,6144,0.05226133267084757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,65536,0.4025599956512451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,6144,0.0740799970097012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,5120,0.06265244219038221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,5120,0.05022044314278496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,4096,0.03996088769700792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,5120,0.06148178047604031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,3584,0.03833955526351929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,4096,0.0564862224790785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,3584,0.055063999361462064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,4096,0.050344000260035195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,3584,0.051334222157796226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,3072,0.035359111097123884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,2560,0.048799110783470996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,3072,0.05176977647675408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,2560,0.032459557056427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,2048,0.02907022171550327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,2560,0.03284622232119242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,1536,0.025449777642885845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,3072,0.03869600097338358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,1024,0.015041778484980265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,1024,0.022410665949185688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,1536,0.042041778564453125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,768,0.011854222251309289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,2048,0.044943110810385815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,1024,0.035777777433395386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,1536,0.021317332983016968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,512,0.008952000074916417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,768,0.02091466718249851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,65536,0.4385040071275499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,768,0.03417955504523383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,2048,0.027109333210521277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,256,0.006064888917737537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,512,0.019225777851210702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,512,0.03225866622394986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,128,0.0038284444146686127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,256,0.027791112661361694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,128,0.015390222271283468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,64,0.0034826666944556763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,256,0.015440889530711703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,32,0.00351555562681622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,128,0.02625244524743822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,64,0.01641155613793267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,32,0.015973332855436537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,16384,0.10701600048277114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,16384,0.095451553662618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,12288,0.0761022236612108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,12288,0.11464622285630967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,16384,0.14830133650037977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,12288,0.08856266736984253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,10240,0.0657857788933648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,10240,0.0953671137491862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,10240,0.07948355542288886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,8192,0.055367999606662326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,65536,0.7443484200371636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,8192,0.07818400197558932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,8192,0.06915199756622314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,7168,0.05077777637375725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,6144,0.045160889625549316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,7168,0.0701999995443556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,6144,0.06106311082839966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,65536,0.32218755616082084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,7168,0.06822044319576688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,6144,0.05891199906667074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,5120,0.04049688908788893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,5120,0.05572444200515747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,4096,0.03675377700063918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,5120,0.04927822285228305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,3584,0.03342399994532267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,4096,0.04074044360054864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,3072,0.031304889255099826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,4096,0.05095555716090732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,3584,0.03590933481852213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,3584,0.049752887752321034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,2560,0.02922933300336202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,2048,0.025945777694384258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,3072,0.031178666485680476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,3072,0.04730755421850416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,2560,0.04449511236614651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,2560,0.026616000466876562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,2048,0.04079466727044847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,1536,0.02453777856296963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,1024,0.012315555579132505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,1536,0.037428445286220975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,1024,0.02141777839925554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,768,0.010036444498433007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,1536,0.01723911033736335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,2048,0.02197866638501485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,768,0.01979200045267741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,1024,0.034528887934154935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,512,0.007697777615653143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,768,0.03329955538113912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,256,0.004699555536111196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,512,0.01647555496957567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,128,0.0036764442920684814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,256,0.015167999598715039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,512,0.031288888719346784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,128,0.015100444356600443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,256,0.02677244444688161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,64,0.0033742222521040174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,128,0.02628177735540602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,64,0.015527110960748462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,32,0.003378666730390655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,65536,0.3808746602800157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,32,0.015455999308162265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,16384,0.0805964469909668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,16384,0.09432622459199692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,12288,0.09095822440253364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,16384,0.11949866347842747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,12288,0.06482488579220243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,12288,0.08599644237094456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,10240,0.056904892126719155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,10240,0.07629422346750896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,10240,0.07234044207466973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,8192,0.048736890157063804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,8192,0.06632444593641493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,7168,0.044734223021401294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,8192,0.06169866853290134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,7168,0.06091022491455078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,7168,0.054620444774627686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,6144,0.04091733362939622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,6144,0.047891553905275136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,6144,0.059791107972462974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,65536,0.570499579111735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,65536,0.32718666394551593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,5120,0.036863111787372164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,65536,0.27560443348354763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,5120,0.0521653327676985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,4096,0.032962666617499456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,4096,0.050444444020589195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,5120,0.04019644525316026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,4096,0.0329093337059021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,3072,0.028904888365003798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,3584,0.030890666776233252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,3072,0.04467466804716322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,2560,0.0262773334980011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,3584,0.04939022329118517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,2560,0.04182222154405382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,3072,0.02566755480236477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,3584,0.030229333374235366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,2048,0.024382222029897902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,1536,0.014171555638313293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,2560,0.021835555632909138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,2048,0.03861155443721347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,1024,0.010448888772063784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,1536,0.023609777291615803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,2048,0.018039110634062026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,1536,0.036706666151682533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,768,0.00869777798652649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,1024,0.020426667398876615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,768,0.01957066688272688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,1024,0.03402311272091336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,768,0.03288088904486762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,512,0.007652444144090016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,512,0.016154666741689045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,256,0.0038648889296584656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,512,0.030390222867329914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,256,0.01495911180973053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,128,0.015086222026083203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,128,0.0035093331502543557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,256,0.02664088871743944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,64,0.003256888853179084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,32,0.0032799999333090256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,128,0.026744888888465032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,64,0.015117333994971381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,32,0.015139554937680563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,16384,0.06604355573654175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,16384,0.08287288745244344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,65536,0.47921424441867405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,16384,0.0935208863682217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,12288,0.07166933351092868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,12288,0.06973689132266574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,12288,0.054302222198910184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,10240,0.04727288749482897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,10240,0.06502844227684869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,8192,0.04131911198298136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,10240,0.06005244122611152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,8192,0.05765689081615872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,8192,0.049110223849614464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,7168,0.03750577900144789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,7168,0.05845777855979072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,6144,0.03418133325046963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,6144,0.03810933232307434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,65536,0.2156604395972358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,7168,0.043291555510626904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,5120,0.03134311238924662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,6144,0.05191822184456719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,4096,0.027874665127860174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,5120,0.04837244417932299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,5120,0.0321084459622701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,3584,0.02602400051222907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,4096,0.026492445005310908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,4096,0.04344800114631653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,3584,0.04416177670160929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,3072,0.023559111687872145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,2560,0.022057778305477563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,3072,0.04114311271243625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,2560,0.03921155465973748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,3584,0.023743111226293776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,3072,0.020855110552575853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,2048,0.020447111792034574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,1536,0.011875555747085147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,2560,0.017886221408843994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,2048,0.037442667616738215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,1536,0.019151111443837483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,1024,0.008900444540712569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,2048,0.014746665954589844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,1536,0.036413331826527916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,1024,0.017350221673647564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,768,0.007518221934636434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,1024,0.03323377834426032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,768,0.015191111299726697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,512,0.005929777605666055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,768,0.032074666685528226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,512,0.014083556003040738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,256,0.0037831109431054858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,512,0.027907556957668726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,256,0.013655111193656921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,128,0.0034479999707804788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,256,0.02643022272321913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,128,0.013382222089502545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,64,0.0031902222997612427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,32,0.0032497776879204642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,128,0.026196445027987163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,64,0.013983110586802164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,32,0.013516444298956128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,65536,0.2785724533928765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,16384,0.06069778071509468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,16384,0.08551377720303005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,16384,0.08319199747509427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,12288,0.049984001451068454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,12288,0.06708888875113593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,12288,0.06507022513283624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,10240,0.044312887721591525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,10240,0.0617848899629381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,65536,0.19935288694169787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,10240,0.05358133382267422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,8192,0.03907910982767741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,65536,0.37842310799492734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,8192,0.05528888768619961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,8192,0.04425333274735344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,7168,0.03553422292073568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,7168,0.038542220989863075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,65536,0.2572106785244412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,6144,0.032396445671717324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,5120,0.03220977716975742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,7168,0.05327644281917148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,4096,0.025850666893853083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,6144,0.05075733198059929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,5120,0.04759200082884895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,3584,0.02420888841152191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,4096,0.04202222161822849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,6144,0.03409511182043288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,3072,0.022544888986481562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,3584,0.042302221059799194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,4096,0.023944000403086346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,5120,0.028920888900756836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,2560,0.021154666940371197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,3584,0.021485333641370136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,3072,0.04173244370354546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,2048,0.013376000026861826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,2560,0.038273778226640486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,2048,0.0199635558658176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,1536,0.011004444625642566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,1536,0.018722666634453666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,2048,0.03633688886960348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,3072,0.018888889087571036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,2560,0.01629333363638984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,1024,0.00833155545923445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,768,0.006866666591829724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,1536,0.034837332036760114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,1024,0.01702933344576094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,1024,0.03289866778585646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,768,0.01479822231663598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,512,0.005751111027267244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,512,0.014067555467287699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,768,0.031331555710898504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,256,0.003580444388919406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,512,0.027232888672086928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,256,0.013487111363146039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,128,0.0033884443756606844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,256,0.026686223016844854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,128,0.013342222405804528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,64,0.003117333269781537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,128,0.025946666797002155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,32,0.0031617778456873367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,64,0.01350933313369751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,32,0.013348444468445249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,16384,0.05827555391523573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,16384,0.07355555560853747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,12288,0.05822044610977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,65536,0.358824888865153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,12288,0.046354668007956616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,12288,0.06236266427569919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,10240,0.041659557157092623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,16384,0.07727289199829102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,10240,0.04829244481192695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,8192,0.0388302206993103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,8192,0.052200890249676175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,10240,0.05810222360822889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,8192,0.040264000495274864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,7168,0.033032000064849854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,7168,0.03535200158754984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,65536,0.1839653386010064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,6144,0.03052622079849243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,7168,0.0506915549437205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,5120,0.027673777606752183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,5120,0.04417422082689074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,6144,0.047296000851525195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,4096,0.024359110328886244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,4096,0.040319111612108015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,5120,0.026240888569090102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,6144,0.030279109875361126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,3584,0.02308533257908291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,3584,0.04053422146373325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,3072,0.021535111798180476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,2560,0.015237333046065437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,3072,0.03897688786188761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,3584,0.019874667127927143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,4096,0.022229333718617756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,2560,0.020871111088328894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,3072,0.017426666286256578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,2048,0.012670222255918713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,2560,0.03729866610633002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,2048,0.019152889649073284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,1536,0.010021333065297868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,1536,0.018166222506099276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,2048,0.03625599874390496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,1024,0.007557333343558841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,1536,0.034550223085615374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,1024,0.015965332587560017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,768,0.006612444503439798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,1024,0.03252622154023912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,768,0.014474666780895658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,512,0.0046977777447965415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,768,0.030982222821977403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,512,0.013979555832015144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,256,0.0035413333939181436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,512,0.027263111538357202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,256,0.013367999758985309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,128,0.0033031110134389666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,128,0.013230221966902414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,256,0.02643644478585985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,64,0.0030666666312350165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,128,0.02603200078010559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,32,0.003104888937539525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,64,0.013447110851605734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,32,0.013249778085284762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,65536,0.22041777769724527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,16384,0.05199199914932251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,16384,0.07107644610934787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,12288,0.04340533415476481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,16384,0.06076888905631172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,12288,0.04684355523851183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,12288,0.05949422385957506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,10240,0.03861866725815667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,10240,0.05583022038141886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,65536,0.1573902236090766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,10240,0.039691554175482854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,8192,0.050386667251586914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,8192,0.03431288732422723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,8192,0.03216000066863166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,7168,0.04924266537030538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,6144,0.028556442923016016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,7168,0.031537777847713895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,6144,0.046430223517947726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,7168,0.02860977914598253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,5120,0.024897777371936377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,65536,0.20191200574239096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,5120,0.04285422298643324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,65536,0.2869306670294868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,4096,0.022954665952258643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,3584,0.016332444217469957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,6144,0.025585777229732935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,4096,0.03899822301334805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,3584,0.02258488867017958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,3584,0.039592888620164655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,5120,0.021839999490314062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,3072,0.014255111416180929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,4096,0.018178666631380718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,3072,0.020543111695183646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,2560,0.012446222205956778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,2560,0.019522666931152344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,3072,0.0380586650636461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,2048,0.010601777997281818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,2048,0.018594667315483093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,2560,0.03681777914365133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,1536,0.008877333667543199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,2048,0.035747554567125105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,1536,0.034176889393064715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,1536,0.017681777477264404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,1024,0.0070568886068132185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,1024,0.0147733340660731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,768,0.006026666611433029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,1024,0.029765334394243028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,768,0.014288889037238227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,512,0.0042257776690853965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,768,0.02826933397187127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,512,0.013920888304710388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,256,0.003469333259595765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,512,0.027152000202072993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,256,0.013284444808959961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,128,0.0032462223122517266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,256,0.026305778159035578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,128,0.013186666700575085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,64,0.0030577776746617425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,32,0.0030746666921509635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,64,0.013180444637934366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,128,0.025785777303907607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,32,0.013134222063753339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,16384,0.04558222161398994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,16384,0.06483378012975057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,65536,0.2392204337649875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,12288,0.03792711098988851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,12288,0.041145778364605375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,16384,0.05188799897829691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,12288,0.055884447362687856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,8192,0.030674666166305542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,10240,0.0339902208911048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,10240,0.05323822299639384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,8192,0.047482666042115956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,10240,0.03526666760444641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,7168,0.02848177817132738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,7168,0.04642311069700453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,6144,0.02584444483121236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,8192,0.029312888781229656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,7168,0.026503110925356548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,6144,0.04342755675315857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,65536,0.1371342208650377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,5120,0.023039999935362074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,5120,0.04051822092798021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,4096,0.02120888895458645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,6144,0.023807111713621352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,3584,0.013366222381591797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,4096,0.037703110112084284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,4096,0.015077332655588785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,5120,0.01850755512714386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,3584,0.02109777761830224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,3072,0.0120328888297081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,3584,0.03869333200984531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,3072,0.019676445258988273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,2560,0.010615111225181155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,2560,0.018805333309703402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,3072,0.03710133168432448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,2048,0.009162666896979014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,2048,0.017810665898852877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,2560,0.03567022085189819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,1536,0.007808888951937358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,2048,0.034840001000298396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,1536,0.016355555918481614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,1024,0.006233777850866318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,1024,0.014648000399271647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,1536,0.03318311108483209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,768,0.004855111241340637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,1024,0.03138755427466498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,512,0.003917333152559069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,768,0.013996443814701505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,512,0.01367377738157908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,768,0.028370665179358587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,256,0.003346666693687439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,512,0.026904000176323786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,256,0.013312000367376538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,128,0.00315022220214208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,256,0.026092444856961567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,128,0.01311377767059538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,64,0.0029440000653266907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,65536,0.1793608930375841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,32,0.003113777687152227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,128,0.025811556312772963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,64,0.013197333448463015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,32,0.013021333350075616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,16384,0.04401955670780606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,16384,0.06380622254477607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,16384,0.046718223227394946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,12288,0.03726933399836222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,12288,0.03668177790111966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,12288,0.054198222027884595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,10240,0.03261155552334256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,10240,0.0318488876024882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,8192,0.02945599953333537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,10240,0.05246666736072964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,8192,0.04698310958014595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,7168,0.02723466687732273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,65536,0.18984533680809867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,6144,0.02442311081621382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,7168,0.04575644599066841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,8192,0.02678666677739885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,5120,0.019048889478047688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,5120,0.022284444835450914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,6144,0.041802667909198336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,7168,0.024280889166726008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,4096,0.01350400017367469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,65536,0.16329333517286512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,6144,0.021799999806616042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,5120,0.04050133294529385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,4096,0.02097333305411869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,65536,0.12265777587890625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,3584,0.012291555603345236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,4096,0.038093331787321306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,3072,0.019215111931165058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,3584,0.020079111059506733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,3072,0.01107377807299296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,3584,0.0384880006313324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,2560,0.009836444424258338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,3072,0.03675644596417745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,2560,0.01826844447188907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,2048,0.00867022242810991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,2560,0.03602577911482917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,2048,0.017768000562985737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,1536,0.007368000017272101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,2048,0.03447644578086005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,1536,0.015160888433456421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,1024,0.005852444304360284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,1024,0.014750222365061441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,1536,0.03196533189879523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,768,0.004209777961174647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,1024,0.028988444142871436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,768,0.01407199932469262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,512,0.003663111064169142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,512,0.026956443985303242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,768,0.028212444649802312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,512,0.013772444592581855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,256,0.00332266671790017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,256,0.013282666603724161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,128,0.0030737777964936364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,128,0.013078221844302284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,64,0.0029502221279674103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,256,0.026378666361172993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,128,0.02632177703910404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,32,0.003012444410059187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,64,0.013080000049538083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,32,0.013088000317414602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,65536,0.17339111698998344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,16384,0.04159288936191135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,16384,0.06190311246448093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,16384,0.043311112456851535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,12288,0.03370399938689338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,12288,0.03458311160405477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,12288,0.053295112318462796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,10240,0.03206844462288751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,8192,0.027419555518362258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,10240,0.050761779149373375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,10240,0.028982222080230713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,8192,0.04596711198488871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,7168,0.025507556067572698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,8192,0.024992000725534227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,7168,0.044343110587861806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,6144,0.02369600037733714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,5120,0.017483555608325534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,6144,0.04169955518510607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,5120,0.022006221943431433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,4096,0.015251555376582675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,7168,0.022504000200165644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,6144,0.01997511088848114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,5120,0.04011911153793335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,4096,0.02055288851261139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,3584,0.011247999966144562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,3584,0.019713777634832595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,3072,0.010288889209429422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,4096,0.03690400057368808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,3072,0.018774222996499803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,3584,0.03783555494414435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,65536,0.12599644396040174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,2560,0.009200889203283522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,3072,0.03624000151952108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,2048,0.007867555651399825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,2560,0.018031999468803406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,2560,0.03544355432192484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,2048,0.0173973325226042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,2048,0.034535109996795654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,1536,0.00676711110605134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,1536,0.015312888556056552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,1024,0.0053119998839166425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,1536,0.031473779016070895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,1024,0.014621333943472968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,768,0.004488888714048597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,1024,0.029192888074451025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,768,0.014078223043017916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,65536,0.15415910879770914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,512,0.003924444317817688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,768,0.028025776147842407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,512,0.013748444616794586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,256,0.0033440000067154565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,512,0.02685955497953627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,256,0.013268444273206921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,128,0.0030871110243929755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,128,0.013066666821638743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,256,0.026368000441127356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,64,0.0029288888391521242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,128,0.025974222355418738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,32,0.002983111060327954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,64,0.013029333617952136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,32,0.012969777815871768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,16384,0.04038666685422262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,12288,0.03332355618476868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,16384,0.05960888995064629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,12288,0.051932444175084434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,12288,0.028055111567179363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,10240,0.030658665630552504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,8192,0.020724443925751578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,10240,0.048828442891438804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,16384,0.035417778624428645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,8192,0.026388444834285315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,10240,0.024620445238219366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,8192,0.04515733321507772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,65536,0.14781867133246526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,7168,0.024619556135601465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,7168,0.043390221065945096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,6144,0.016923555069499545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,6144,0.02374222212367588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,5120,0.014882667197121514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,6144,0.04123377799987793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,7168,0.018739556272824604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,5120,0.021617778473430212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,4096,0.013058666553762225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,5120,0.03908355699645148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,4096,0.020522667302025687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,3584,0.012180444267061023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,3584,0.019374221563339233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,3072,0.011370666325092316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,4096,0.03724177678426107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,3584,0.036630223194758095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,3072,0.018589332699775696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,2560,0.00850044439236323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,2560,0.017934223016103108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,3072,0.03610666592915853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,2048,0.0074417781498697065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,2560,0.03513155712021722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,2048,0.015861334072219003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,1536,0.006528000036875407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,2048,0.03413955701722039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,65536,0.10691910982131958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,1536,0.015278221832381355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,1024,0.004592000196377437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,1536,0.03059200114674038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,1024,0.014576888746685453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,1024,0.02881511052449544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,768,0.004374222209056218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,768,0.013947554760509066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,512,0.003689777933888965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,512,0.013630222115251752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,768,0.02784088916248745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,65536,0.14594577418433294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,256,0.003305777700410949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,512,0.02697155541843838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,256,0.013302221894264221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,128,0.003059555465976397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,256,0.026391999589072332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,128,0.013019555144839816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,64,0.0029377777957253983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,128,0.025565332836574976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,32,0.0029644444584846497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,64,0.013053333593739403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,32,0.013239999612172445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,16384,0.03840177920129564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,16384,0.0581208864847819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,16384,0.02922933300336202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,12288,0.028953777419196233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,12288,0.05122488737106323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,10240,0.028575999869240656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,65536,0.1248977714114719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,12288,0.02368533280160692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,10240,0.04805155595143636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,8192,0.025231111380789015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,10240,0.02075822154680888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,7168,0.015838222371207345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,7168,0.024138665861553613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,8192,0.04315911067856682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,6144,0.0144533332851198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,8192,0.017255110873116385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,7168,0.0429111123085022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,6144,0.02255022194650438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,65536,0.11285066604614258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,5120,0.012910222013791403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,65536,0.14915555053287083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,6144,0.04056888818740845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,5120,0.021009777983029682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,4096,0.01147911118136512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,4096,0.0199626667631997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,5120,0.03838310970200433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,3584,0.010822222464614444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,4096,0.03674577673276266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,3584,0.01906133360332913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,3072,0.010229333407349056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,3072,0.018248889181349013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,3584,0.036743111080593534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,2560,0.00774133370982276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,2560,0.01648177703221639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,3072,0.035447110732396446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,2048,0.0069617778062820435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,2560,0.03275110986497667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,2048,0.015777778294351365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,1536,0.00573600000805325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,2048,0.031621333625581526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,1536,0.015065777632925244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,1024,0.004256000121434529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,1536,0.03047377864519755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,1024,0.014426666829321118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,768,0.0038888889054457345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,1024,0.029000888268152874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,768,0.013929777675204806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,512,0.004106666478845808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,512,0.02681244413057963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,768,0.02776977750990126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,512,0.013617777989970313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,256,0.003212444484233856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,256,0.013383111192120446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,128,0.0030382221771611106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,32,0.0029031110720502008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,256,0.02595377796226078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,128,0.013074666261672974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,64,0.0029200000895394217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,128,0.025721778472264607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,64,0.012904889053768583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,32,0.012877333495352002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,65536,0.10093155834409927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,16384,0.0358497765329149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,12288,0.029885331789652508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,16384,0.05695199966430664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,12288,0.04974577824274699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,10240,0.017279111676745944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,10240,0.027303111222055223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,8192,0.014630221658282809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,10240,0.04698933495415581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,12288,0.019215111931165058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,8192,0.0244186669588089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,16384,0.023430221610599097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,7168,0.013425778183672162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,7168,0.0232595553000768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,8192,0.043266667260064014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,6144,0.012545777691735161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,7168,0.04243910974926419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,6144,0.02180444366402096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,5120,0.011412444214026133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,6144,0.04011288947529263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,5120,0.019705777366956074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,4096,0.010289777484205034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,5120,0.038988444540235735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,3584,0.00978311151266098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,4096,0.018713777263959248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,3584,0.018086221483018663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,4096,0.03635377685228983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,3072,0.009372444616423713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,3584,0.0351475543446011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,2560,0.0073520003093613525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,3072,0.0166302224000295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,3072,0.03340889016787211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,65536,0.1030622190899319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,65536,0.13875199688805476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,2560,0.01589866644806332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,2048,0.015068444940778943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,2048,0.0063893331421746155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,2560,0.03338133295377096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,1536,0.00490044429898262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,2048,0.03219022353490194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,1024,0.004259555704063839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,1536,0.014829332629839579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,1536,0.030404445197847154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,768,0.003903111235962974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,768,0.013921777407328287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,1024,0.014056000444624158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,1024,0.028704888290829126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,512,0.013516444298956128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,512,0.0035911111368073356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,768,0.027679110566775005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,256,0.003269333392381668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,512,0.02695200012789832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,128,0.003051555612021022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,256,0.01314399970902337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,256,0.025826666090223525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,128,0.013074666261672974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,64,0.002891555635465516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,32,0.0028986665937635633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,128,0.025468443830808003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,64,0.013077333569526672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,32,0.012814222110642327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,65536,0.07612178060743544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,16384,0.026068444053332012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,12288,0.01484444406297472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,16384,0.047967112726635404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,12288,0.022877333892716303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,12288,0.04436622063318888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,10240,0.01334755536582735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,10240,0.02142488956451416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,16384,0.01774488886197408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,8192,0.011628444823953839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,8192,0.019912888606389362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,10240,0.04245333207978142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,7168,0.01107022249036365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,8192,0.03984533415900336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,7168,0.019135110908084445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,6144,0.009879111415810054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,7168,0.039424889617496066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,6144,0.01847466660870446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,5120,0.009665778113736046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,5120,0.01809155609872606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,65536,0.06623911195331149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,6144,0.03794577717781067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,4096,0.00906666699382994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,5120,0.03619111246532864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,65536,0.12077332867516412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,4096,0.0158906661801868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,3584,0.008889777792824639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,3584,0.03272355596224467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,4096,0.03241066800223456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,3584,0.01608711149957445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,3072,0.00813688917292489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,3072,0.015487111277050443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,2560,0.006271111054552927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,3072,0.03202222122086419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,2560,0.015024888846609326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,65536,0.05526577764087253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,2048,0.005642666584915585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,2560,0.031517333454555936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,2048,0.014694222145610385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,1536,0.02976977825164795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,1536,0.004855999930037392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,2048,0.030598223209381104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,1536,0.014293332894643148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,1024,0.004217777815130022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,1024,0.01423288881778717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,768,0.0038488888078265717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,768,0.013613333304723104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,1024,0.028548445966508653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,512,0.0035137778355015647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,768,0.027853333287768896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,512,0.013379555609491138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,256,0.0032231110251612137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,512,0.026576000783178542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,256,0.013040000365840064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,128,0.00300177786913183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,256,0.025781333446502686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,128,0.012838222086429596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,64,0.0029013332807355454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,128,0.025632888078689575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,32,0.0028311111446883944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,64,0.012687111066447364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,32,0.012684444586435953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,16384,0.014797333214018079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,16384,0.02485688858562046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,12288,0.013056000073750814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,12288,0.0222053329149882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,16384,0.0473217765490214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,10240,0.011749332977665795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,12288,0.04281333420011732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,10240,0.021003555920388963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,8192,0.010572444233629437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,65536,0.06332444482379489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,10240,0.04175466630193922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,8192,0.019537778364287484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,7168,0.009919111099508073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,6144,0.009039999710188972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,7168,0.018957333432303537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,8192,0.0384115543630388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,6144,0.017302221722073026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,7168,0.038329776790406965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,5120,0.008993777963850234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,65536,0.11189422342512344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,6144,0.036496887604395546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,5120,0.016353777713245816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,4096,0.008417777717113495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,5120,0.03422666589419047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,4096,0.01608088943693373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,3584,0.007992000215583378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,3072,0.007499555746714274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,3584,0.015659555792808533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,4096,0.03196533189879523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,65536,0.04504088891877068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,3584,0.03262844350602891
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,3072,0.015409777561823526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,2560,0.006567111031876669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,2560,0.014879110786649915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,3072,0.03156355685657925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,2048,0.005509333478079901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,2560,0.0314026673634847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,2048,0.01479644411140018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,1536,0.004907555464241239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,2048,0.030272887812720403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,1536,0.014162666267818876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,1024,0.0041742221348815495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,768,0.0038355555799272326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,1536,0.02951644526587592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,1024,0.014102222190962898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,768,0.013711111413107978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,1024,0.028441776831944782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,512,0.0034924443397257063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,768,0.027708444330427382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,512,0.013354666531085968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,256,0.0031973332580592898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,512,0.026670222481091816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,256,0.013252444565296173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,128,0.0029964444951878656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,256,0.02602400051222907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,128,0.012792888614866467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,64,0.002834666727317704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,128,0.025615110993385315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,32,0.0028284444577164124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,64,0.012706666356987424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,32,0.012960889273219638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,16384,0.02461244497034285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,16384,0.01184622198343277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,12288,0.010303999814722273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,16384,0.04749333196216159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,12288,0.022131555610232886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,10240,0.00981244444847107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,12288,0.04302488764127096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,10240,0.019852444529533386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,8192,0.009336888790130615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,65536,0.06061066521538628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,10240,0.04166311025619507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,8192,0.01800622211562263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,65536,0.11237511369917129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,7168,0.008646222452322641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,8192,0.03618933426009284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,7168,0.01754755609565311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,6144,0.008221333225568136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,65536,0.032052444087134466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,6144,0.034747554196251765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,7168,0.03637511200375027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,6144,0.016693333784739178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,5120,0.007877333296669854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,5120,0.016190222567982145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,4096,0.007161777880456712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,5120,0.033328887489106916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,3584,0.007680888805124495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,4096,0.015767110718621146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,4096,0.03141511148876614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,3584,0.015258666541841296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,3072,0.007125332951545715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,3584,0.032425777779685125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,3072,0.015030221806632148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,2560,0.006236444330877728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,3072,0.032130665249294706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,2560,0.01457333399189843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,2048,0.005519111123349931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,2560,0.031168000565634832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,2048,0.01445777714252472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,1536,0.004870222260554631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,2048,0.030592888593673706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,1536,0.014302222265137566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,1024,0.0041546668443414895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,1536,0.029069334268569946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,1024,0.013765333427323235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,768,0.003867555409669876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,768,0.027956446011861164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,1024,0.028007109959920246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,768,0.013439111411571503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,512,0.0035031110876136353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,512,0.013199999928474426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,256,0.0032026666320032547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,256,0.013268444273206921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,512,0.026337777574857075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,128,0.0029813332690132987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,256,0.025723555021815833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,128,0.012823111481136747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,64,0.002830222249031067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,128,0.025479111406538222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,32,0.0028275555620590844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,64,0.012845333251688214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,32,0.012508444488048553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,16384,0.008889777792824639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,65536,0.058456891112857394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,16384,0.023636443747414485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,12288,0.008380444513426887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,65536,0.1134951114654541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,12288,0.02082577844460805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,16384,0.0469706654548645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,10240,0.007829333345095316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,12288,0.04181955589188469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,65536,0.021207999851968553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,10240,0.0192239996459749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,8192,0.007515555454625024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,10240,0.03932177689340379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,8192,0.018152889278199937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,7168,0.007107555866241455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,8192,0.03603200117746989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,7168,0.017320000463061862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,6144,0.0070293330483966405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,7168,0.035818666219711304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,5120,0.007230222225189209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,6144,0.016807110773192514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,5120,0.03324266605907016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,6144,0.034782220919926964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,5120,0.016183111402723525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,4096,0.0313351121213701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,4096,0.006778666542636023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,4096,0.015599111715952555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,3584,0.007295110987292395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,3072,0.006978666616810693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,3584,0.015474667151769003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,2560,0.006151999864313338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,3072,0.014999111493428549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,2560,0.031119111511442397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,3584,0.032334221733940974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,2560,0.014579556054539151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,3072,0.0315324432320065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,2048,0.00545155546731419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,2048,0.014520888527234396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,1536,0.0047866664826869965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,1536,0.01424800025092231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,2048,0.030239999294281006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,1024,0.004156444635656145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,1536,0.029234667619069416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,1024,0.013802666631009845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,768,0.0038168889780839286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,768,0.02767555581198798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,768,0.013507555756303998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,1024,0.028057777219348486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,512,0.003526222374704149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,512,0.013223999904261695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,256,0.003108444520168834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,512,0.026199999782774184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,256,0.013005333642164866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,128,0.002987555538614591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,256,0.026375111606385972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,128,0.012861332959598966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,64,0.0028248888750871024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,32,0.0028373334142896864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,128,0.02537333303027683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,64,0.012647111382749347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,32,0.012479110724396177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,65536,0.014535999960369535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,16384,0.007299555672539606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,65536,0.05739021963543362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,16384,0.022801778382725183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,12288,0.006963555183675554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,65536,0.11136444409688313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,16384,0.04469866553942362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,12288,0.02051911089155409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,10240,0.006931555353932911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,12288,0.040334221389558576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,10240,0.019164444671736825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,10240,0.038907554414537214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,8192,0.006894222150246303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,8192,0.017921777235137094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,7168,0.00675644435816341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,8192,0.035800000031789146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,7168,0.017239999439981248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,6144,0.006694222076071634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,7168,0.03639377819167243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,6144,0.016551999582184684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,5120,0.006959999601046245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,6144,0.03469244307941861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,5120,0.01608444419172075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,4096,0.006574222197135289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,5120,0.03314844436115689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,4096,0.01573244399494595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,3584,0.0070293330483966405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,4096,0.03162577748298645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,3584,0.015252444479200574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,3072,0.006726222319735422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,3584,0.032433778047561646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,3072,0.03204800022972955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,2560,0.014576888746685453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,3072,0.014969777729776172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,2560,0.006108444597986009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,2048,0.005444444302055571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,1536,0.004810666872395409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,2048,0.014521777629852295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,2560,0.030930668115615845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,1536,0.014083556003040738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,1024,0.004098666624890434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,2048,0.030157334274715845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,1024,0.027967111931906805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,1024,0.013639111485746173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,1536,0.028866665230857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,768,0.0037813331517908308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,768,0.013491555220550962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,512,0.003491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,768,0.027152889304690894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,512,0.013270222478442721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,256,0.00309955556359556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,512,0.02646666599644555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,256,0.013002666334311167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,128,0.0029528888149393927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,256,0.025803556044896443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,128,0.012822222378518848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,64,0.0027706666539112725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,32,0.0028106667515304354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,64,0.012474666866991254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,32,0.012895111408498553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,65536,0.012856889102194043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,128,0.02557777696185642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,16384,0.006878222028414409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,65536,0.05671999851862589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,12288,0.006893333461549547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,16384,0.02279111080699497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,10240,0.0069271110826068455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,12288,0.02032444377740224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,8192,0.006847111301289664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,10240,0.019080888893869188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,8192,0.01777866648303138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,7168,0.006636444479227066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,6144,0.006506666541099548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,7168,0.01719377769364251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,6144,0.016527111331621807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,5120,0.00700711127784517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,4096,0.0064533332155810455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,5120,0.016036444240146212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,3584,0.0069715554515520734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,4096,0.015477332803938123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,3072,0.006627555522653792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,3584,0.015213333898120456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,2560,0.006119110931952794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,3072,0.014879110786649915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,2048,0.005480000128348668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,2560,0.014627556006113688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,1536,0.004755555755562252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,2048,0.01424711114830441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,1024,0.004100444416205089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,1536,0.014057776994175382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,768,0.003689777933888965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,1024,0.013930666777822705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,512,0.003432000055909157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,768,0.013417777915795645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,256,0.003031111011902491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,512,0.013120888835854001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,128,0.002962666667169995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,256,0.012867555850081973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,64,0.0028328889360030494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,128,0.012684444586435953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,32,0.0027742222365405825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,64,0.012436444560686747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,65536,0.012345777617560493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,32,0.012330667012267642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,16384,0.007174222005738153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,65536,0.05598666932847765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,12288,0.007116444408893585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,16384,0.02294844388961792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,12288,0.02055288851261139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,10240,0.006948444578382704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,8192,0.006698666761318843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,10240,0.018956444329685636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,7168,0.006581333362393909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,8192,0.017838221457269456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,7168,0.017349332571029663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,6144,0.006522666662931442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,5120,0.006836444553401735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,6144,0.01664888858795166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,4096,0.0064195555945237475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,5120,0.015941333439615037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,3584,0.0070257774657673305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,4096,0.015415110521846347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,3072,0.006573333508438534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,3584,0.015341333217091031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,3072,0.014861333701345654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,2560,0.006041777630647023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,2560,0.014459555347760519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,2048,0.005491555564933353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,1536,0.004744889007674323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,2048,0.014312000738249885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,1024,0.0041128889554076726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,1536,0.014112888111008538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,768,0.0037173334923055437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,1024,0.013605333036846586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,512,0.003415999934077263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,768,0.013621333572599622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,256,0.0030497778207063675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,512,0.013229333692126803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,128,0.002871111035346985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,256,0.012871999707486896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,64,0.0027795556104845474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,128,0.012694222231705984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,64,0.012715555727481842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,32,0.0027235555979940626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,32,0.0123831108212471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,12288,0.3521146774291992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,12288,0.2855315473344591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,10240,0.23810577392578125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,16384,0.372545772128635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,16384,0.41982221603393555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,10240,0.47454579671223956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,12288,0.5678755442301432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,8192,0.19467555152045357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,10240,0.29854488372802734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,16384,0.7483270963033041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,8192,0.24690755208333334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,7168,0.19714044200049508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,6144,0.15061955981784397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,7168,0.22048444218105742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,8192,0.38163288434346515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,6144,0.19392800331115723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,5120,0.1287884447309706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,5120,0.1622577773200141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,6144,0.2860657903883192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,4096,0.12007466952006023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,7168,0.3311279879675971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,4096,0.1922017733256022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,3584,0.09599466456307305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,4096,0.12953955597347683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,5120,0.2381866772969564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,3072,0.09438666370179917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,3584,0.12432532840304905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,3072,0.10951822333865696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,2560,0.07406400309668647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,3072,0.14582133293151855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,2560,0.09972000122070312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,2048,0.06268266836802165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,3584,0.16868889331817627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,2560,0.12204089429643418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,2048,0.0819715526368883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,1536,0.07536622550752428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,1024,0.04186577929390801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,1024,0.05488533443874783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,768,0.03788355655140347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,1536,0.05244622296757168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,2048,0.09796977705425686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,1536,0.06710133287641737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,768,0.04020266731580099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,256,0.02911111050181919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,512,0.03382133444150289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,256,0.035944888989130654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,1024,0.05144800080193413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,128,0.009413333402739631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,256,0.015710221396552194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,128,0.026710222164789837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,64,0.00665866666369968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,512,0.028145776854621038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,128,0.03277600142690871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,512,0.04332000017166138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,768,0.04872533347871569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,32,0.0047484445903036325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,64,0.023896889554129705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,32,0.024023110667864483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,12288,0.08485066890716553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,16384,0.10737422439787124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,16384,0.14020800590515137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,12288,0.11146488454606797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,12288,0.13946310679117838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,10240,0.08260889185799493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,16384,0.18534578217400444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,10240,0.09971111350589329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,8192,0.06202399730682373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,10240,0.11631911330752903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,8192,0.08619999885559082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,8192,0.09472355577680801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,7168,0.05590666664971245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,7168,0.07698666387134127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,6144,0.050660444630516895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,6144,0.06922488742404513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,65536,0.3784160084194607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,7168,0.0829191141658359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,6144,0.07255644268459745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,5120,0.04533511069085863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,5120,0.06266133652793036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,4096,0.04000533289379544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,5120,0.06043022208743625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,4096,0.057448003027174205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,3584,0.03701688845952352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,4096,0.05684266487757365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,3584,0.0539608895778656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,3072,0.034436444441477455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,3584,0.043728000587887235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,3072,0.05003199974695841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,3072,0.03801777627733018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,2560,0.03158488869667053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,2560,0.04998844530847338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,2560,0.03244266576237149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,2048,0.028394665982988145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,1536,0.024991111622916326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,2048,0.026497777965333726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,1536,0.03922755519549052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,65536,0.4315484364827474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,1024,0.021949332621362474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,768,0.011766222616036734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,2048,0.043266667260064014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,1024,0.014781332678265043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,1024,0.03491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,512,0.008861333131790161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,768,0.033097777101728655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,256,0.005912000106440649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,512,0.018918222851223417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,1536,0.02090577781200409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,128,0.003624000069167879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,256,0.014858666393491956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,512,0.031381332212024264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,64,0.0033662221911880705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,256,0.027124444643656414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,128,0.014450665977266101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,768,0.020971554848882887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,128,0.026354667213228013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,32,0.003370666669474708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,64,0.015070221490330167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,32,0.015272888872358533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,16384,0.08608888917499119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,16384,0.11144977145724827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,12288,0.06873244709438749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,12288,0.11273954974280463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,16384,0.1466497712665134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,12288,0.09110755390591091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,10240,0.060159113672044545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,10240,0.0890808900197347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,8192,0.051797333690855235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,10240,0.08188888761732313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,8192,0.06899555524190266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,8192,0.07250489128960504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,7168,0.04677244358592563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,7168,0.06615644693374634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,7168,0.06514666477839152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,6144,0.04683644572893778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,65536,0.2948391172620985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,6144,0.06058400207095676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,6144,0.055200000603993736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,5120,0.03844177722930908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,5120,0.0596382220586141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,4096,0.034407999780442976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,4096,0.03713066710366143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,5120,0.047785778840382896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,65536,0.3662666744656033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,4096,0.05099733339415657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,3584,0.03222044308980306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,3584,0.04858577913708157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,3072,0.03182222114668952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,2560,0.027653333213594224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,3072,0.04598933458328247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,3584,0.03312888741493225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,2560,0.04314844475852119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,3072,0.030219554901123047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,2048,0.024903110331959192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,65536,0.7440969149271647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,2048,0.038909332619773015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,1024,0.011794666449228922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,1536,0.022856889499558344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,2048,0.02067466742462582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,1536,0.03647999962170919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,2560,0.02516799999607934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,1024,0.02057066725360023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,768,0.009945777555306753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,1024,0.03291733397377862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,1536,0.017102221647898357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,768,0.02017333275742001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,512,0.007744000189834171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,768,0.03199999862247043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,512,0.01663466625743442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,256,0.004591111093759537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,512,0.030789332257376775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,128,0.003645333151022593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,256,0.014494222071435718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,128,0.014467555615637036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,256,0.02659911082850562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,64,0.0032879999942249725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,128,0.026218665970696345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,32,0.003318222239613533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,64,0.014236445228258768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,32,0.015014222926563688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,16384,0.07658577627605863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,16384,0.09242133299509685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,65536,0.5647030936347114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,16384,0.11698577139112686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,12288,0.06093777550591362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,10240,0.053284443087048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,12288,0.09207466575834487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,12288,0.07903022236294217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,10240,0.07447110944324069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,10240,0.07074311044481066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,8192,0.04632177617814806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,8192,0.06252088811662462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,8192,0.060518225034077965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,7168,0.042564445071750216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,7168,0.06048266755210029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,7168,0.05348799957169426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,6144,0.038915554682413735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,5120,0.03828977876239353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,6144,0.056221332814958364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,6144,0.04675199919276767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,5120,0.0521395570702023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,5120,0.04082489013671875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,4096,0.031832890378104314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,65536,0.25969600677490234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,4096,0.047938668065600924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,3584,0.028598222467634413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,4096,0.03215377860599094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,3072,0.02801333202256097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,65536,0.32117509841918945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,2560,0.02571288910177019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,3072,0.043318221966425575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,3584,0.030052443345387776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,3584,0.04632177617814806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,2048,0.02360711163944668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,2560,0.04228711128234863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,2048,0.03808177842034234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,1536,0.013964444398880005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,1536,0.021776000658671062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,2560,0.021496888664033677
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,1024,0.01053600013256073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,1536,0.035282668140199445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,3072,0.02509422269132402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,768,0.008649778034951951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,1024,0.020119110743204754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,768,0.01885688967174954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,1024,0.032885332902272545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,512,0.006837333242098491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,2048,0.017719111508793302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,768,0.03155555658870273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,512,0.015328889091809591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,256,0.0037973332736227247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,512,0.027942223681343928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,256,0.014450665977266101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,128,0.003375111147761345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,128,0.026312889324294195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,256,0.0265964451763365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,128,0.014132445057233175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,64,0.003154666680428717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,32,0.0031439999325407874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,64,0.01421866648726993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,32,0.014740443891949125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,16384,0.06278133392333984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,16384,0.08058844672309028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,12288,0.05138577686415779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,12288,0.07106577687793307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,16384,0.09325510925716823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,12288,0.06991644700368245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,10240,0.04520977867974175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,10240,0.059728887346055776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,8192,0.03976088762283325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,10240,0.06318844689263238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,8192,0.05638666947682699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,8192,0.048383110099368624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,7168,0.03657244311438666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,7168,0.055220444997151695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,7168,0.0427706672085656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,6144,0.03347377644644843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,65536,0.47633955213758683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,65536,0.20171111159854463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,6144,0.05145244465933906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,6144,0.03770133190684848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,65536,0.2677564356062147
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,5120,0.0306675566567315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,5120,0.047648890150917896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,4096,0.027296000056796606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,3584,0.027338667048348322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,5120,0.03199022346072727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,3584,0.04247911108864678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,4096,0.04473155405786303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,3072,0.023339556323157415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,4096,0.026125333375400964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,3072,0.039834668238957725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,2560,0.0217866665787167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,3584,0.023393778337372675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,2560,0.03790844480196635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,2048,0.020305777589480083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,1536,0.011610666910807291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,1536,0.018782221608691745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,2560,0.020087111327383254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,2048,0.03628977802064683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,2048,0.01443733274936676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,3072,0.02052622205681271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,1024,0.008788444101810455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,1024,0.017204445269372728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,1536,0.03401688734690348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,1024,0.0322675539387597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,768,0.0074106670088238185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,768,0.01425866617096795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,512,0.006010666489601135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,768,0.031015998787350122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,512,0.013589333328935834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,256,0.0036515556275844574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,512,0.02843022346496582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,256,0.013165333204799227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,128,0.0032986667421129015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,256,0.02644088864326477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,128,0.012996444271670448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,64,0.0031591111587153543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,128,0.025944888591766357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,64,0.0129466669427024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,32,0.0031795555518733132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,32,0.013543999857372709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,16384,0.05830044216579861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,16384,0.07727022303475274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,65536,0.3764435450236003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,12288,0.06670933299594455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,12288,0.04775111211670769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,12288,0.06628355715009901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,16384,0.08593155278099908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,10240,0.042232887612448804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,10240,0.06483822398715548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,8192,0.0369102226363288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,10240,0.06184799803627861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,8192,0.04304533203442892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,8192,0.05521688858668009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,7168,0.033831109603246055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,7168,0.03817333446608649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,7168,0.053745776414871216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,6144,0.03154222170511881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,6144,0.033590220742755465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,6144,0.049694223536385425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,5120,0.028991109795040552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,4096,0.025633777181307476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,5120,0.028581334484948054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,4096,0.0432933337158627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,3584,0.023546665906906128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,65536,0.24457777871025932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,5120,0.04692622356944614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,4096,0.02347999976740943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,3584,0.042075554529825844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,65536,0.17938221825493705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,3072,0.02370400064521366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,3072,0.03880177934964498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,2560,0.021106666988796655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,2048,0.01329955541425281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,3072,0.018725333942307364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,3584,0.021199110481474135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,2560,0.03764177693261041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,2048,0.019596444235907663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,1536,0.01090400003724628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,2560,0.016024000114864774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,2048,0.035643554396099515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,1536,0.018024888303544786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,1024,0.00810933361450831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,1024,0.01682222220632765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,1536,0.03394666645261977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,768,0.007045333584149678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,1024,0.031515555249320135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,768,0.014899555179807874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,768,0.030093332131703694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,512,0.006091555373536215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,512,0.013579555683665805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,256,0.003510222252872255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,512,0.027227555712064106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,256,0.013050666285885705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,128,0.0032311110860771606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,256,0.02642400066057841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,128,0.013040000365840064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,64,0.0030133333057165146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,128,0.025948445002237957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,32,0.003007111036115222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,64,0.012864000267452665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,32,0.013586666848924426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,16384,0.05335466729270088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,16384,0.07363288932376437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,12288,0.0439226660463545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,16384,0.07604710923300849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,12288,0.06279289060168795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,12288,0.05728889173931546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,10240,0.03939911060863071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,10240,0.047488888104756675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,10240,0.05788711044523451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,8192,0.03494577937655979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,8192,0.051628443929884166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,65536,0.32279109954833984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,8192,0.03971199856864081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,7168,0.03251733382542928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,7168,0.034712000025643244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,7168,0.05048177639643351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,65536,0.16205867131551108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,65536,0.21050577693515352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,5120,0.027064889669418335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,6144,0.02994222111172146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,4096,0.023689778314696416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,5120,0.044723557101355664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,6144,0.04811911119355095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,6144,0.0347653329372406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,4096,0.041606220934126116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,3584,0.02273955610063341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,5120,0.025773333178626165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,3584,0.03979644510481093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,4096,0.021740444832377966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,3072,0.02134399943881565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,2560,0.014935111006100973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,2560,0.020390222469965618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,3072,0.038147555457221136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,3584,0.019484443797005545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,2048,0.012658666405412885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,2048,0.018990221950742934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,1536,0.010046222143703038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,2560,0.03711822297837999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,1536,0.017822222577200997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,2048,0.03491377830505371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,1024,0.007629333270920648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,1024,0.015983111328548856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,768,0.006548444430033366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,1536,0.03346577617857192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,1024,0.031195554468366835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,768,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,512,0.0047600000268883174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,768,0.027694221999910142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,3072,0.017143110434214275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,512,0.01349511080318027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,256,0.0034471110751231513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,512,0.027129777603679236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,256,0.013120888835854001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,128,0.0031928889867332247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,256,0.026779555612140234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,128,0.012860444684823355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,64,0.002987555538614591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,32,0.013391999734772576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,128,0.026126222478018865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,32,0.003032888803217146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,64,0.012886222038004132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,16384,0.04773066772354973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,16384,0.07044088840484619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,12288,0.04639200038380093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,12288,0.039812445640563965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,16384,0.06040355232026842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,10240,0.035848001639048256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,12288,0.060441778765784375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,10240,0.05628444751103719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,8192,0.032073779238594904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,10240,0.03922577699025472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,65536,0.28393154674106175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,8192,0.05059555504057142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,8192,0.03194400005870395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,6144,0.02734400000837114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,7168,0.029735998974906072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,6144,0.046543111403783165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,7168,0.04916622241338094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,6144,0.025233777032958135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,5120,0.024303111765119765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,7168,0.028374221589830186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,5120,0.04500444398985969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,4096,0.022329777479171753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,65536,0.14257066779666477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,3584,0.021393777595625982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,4096,0.040384001202053495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,4096,0.017834666702482436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,65536,0.18772978252834746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,3072,0.014177777700954013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,5120,0.02156266735659705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,3072,0.020382222202089097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,3584,0.03870222303602431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,2560,0.012346666720178394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,2560,0.019415999452273052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,3584,0.016055110428068373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,3072,0.03730577892727322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,2048,0.010494222243626913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,2560,0.03633066680696275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,2048,0.018183110488785636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,1536,0.008828444613350762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,2048,0.034443553951051496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,1536,0.017332444588343304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,1024,0.0069013333155049225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,1536,0.03311555584271749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,1024,0.01401422255569034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,1024,0.030390222867329914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,768,0.00608711110221015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,768,0.013784888717863293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,512,0.004319111092223061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,768,0.027544000082545813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,512,0.013437333206335703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,256,0.0033368888414568375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,512,0.027013333307372198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,256,0.01315199997689989
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,128,0.0031884445084465873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,256,0.02608355548646715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,128,0.012798222402731577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,64,0.002969777832428614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,128,0.02584533393383026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,32,0.0029768887907266617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,64,0.012763555679056378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,32,0.013358222113715278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,16384,0.0425182216697269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,16384,0.06562488608890109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,12288,0.035191112094455294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,16384,0.05150844322310554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,12288,0.05779733260472616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,12288,0.04078844520780776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,10240,0.031924443112479314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,10240,0.03459999958674113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,65536,0.1196204423904419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,65536,0.2340595457288954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,10240,0.0537066658337911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,8192,0.02862311071819729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,7168,0.028390222125583228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,8192,0.04773155517048306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,7168,0.047184000412623085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,6144,0.02410488824049632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,8192,0.02884977724817064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,5120,0.02235466738541921
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,6144,0.04339289002948337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,7168,0.025966222087542217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,5120,0.04214400053024292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,4096,0.020820443828900654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,65536,0.17474844720628527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,6144,0.023388443721665278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,4096,0.03922577699025472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,3584,0.015572445260153877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,3584,0.020251555575264823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,3072,0.012074666718641916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,5120,0.01815111107296414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,3584,0.03804622093836466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,3072,0.019186667270130582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,2560,0.010585777461528778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,4096,0.01717155509524875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,3072,0.036716444624794856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,2560,0.0184888889392217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,2048,0.009012444151772393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,2048,0.0177208897140291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,2560,0.03570488757557339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,1536,0.00777688870827357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,2048,0.0340008901225196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,1536,0.01627200014061398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,1024,0.006288888967699475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,1024,0.013787555197874704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,1536,0.03265955713060167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,768,0.004825777891609403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,1024,0.028121779362360638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,768,0.01349244432316886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,512,0.003760888758632872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,512,0.02695466743575202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,768,0.027645332945717707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,256,0.0033937777496046494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,512,0.013350221845838757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,256,0.012975999878512489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,128,0.003087999920050303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,256,0.026047110557556152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,128,0.012727110750145383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,64,0.002936888900068071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,128,0.026047999660174053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,64,0.012615111139085559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,32,0.003056888985964987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,32,0.013250666360060373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,16384,0.03891466723548041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,16384,0.06562399864196777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,12288,0.03594311078389486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,12288,0.03309866786003113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,10240,0.03004266818364461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,16384,0.04552177919281853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,65536,0.18750222524007162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,10240,0.0528053343296051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,12288,0.05616622500949436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,10240,0.031195554468366835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,8192,0.028414222929212783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,8192,0.04681422313054403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,7168,0.024551999237802293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,7168,0.04551377892494202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,6144,0.023056000471115112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,8192,0.026084444589085046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,6144,0.04256800148222181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,6144,0.021367111139827307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,7168,0.02384000023206075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,5120,0.02148355543613434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,4096,0.01568711135122511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,5120,0.041095998552110456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,4096,0.020259555843141343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,5120,0.01866311166021559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,3584,0.014295111099878946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,4096,0.03898577888806661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,3584,0.01944177846113841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,65536,0.1066968904601203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,3584,0.037992000579833984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,3072,0.011095110740926532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,3072,0.018958222534921434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,2560,0.009840889109505547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,3072,0.03621866636806064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,2560,0.018387556076049805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,2048,0.00850666645500395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,2048,0.017224889662530687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,2560,0.03530399998029073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,1536,0.007375111182530721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,2048,0.03374044431580438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,1024,0.005884444548024072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,1536,0.015572445260153877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,768,0.004775111046102312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,1024,0.013925333817799887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,1536,0.03235200047492981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,768,0.013577777478430005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,1024,0.027816888358857896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,512,0.0036862223512596558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,768,0.027227555712064106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,256,0.003305777700410949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,512,0.013377777404255338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,512,0.02680177821053399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,65536,0.16543377770317927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,128,0.0030533334033356774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,256,0.013073777986897362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,256,0.025920000341203477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,128,0.012710221939616732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,64,0.0029519999192820657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,128,0.02587288949224684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,64,0.012713777522246042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,32,0.002991111121243901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,32,0.013344888885815939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,65536,0.171834667523702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,16384,0.03573777940538194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,16384,0.0626737806532118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,12288,0.03260622090763516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,12288,0.032095111078686185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,16384,0.04072177741262648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,10240,0.027270222703615826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,12288,0.055946667989095054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,10240,0.02844888965288798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,8192,0.02416977783044179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,10240,0.05202577842606438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,7168,0.02271911170747545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,8192,0.045999109745025635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,7168,0.04528977804713779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,6144,0.02198311189810435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,5120,0.016903999778959487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,7168,0.022016889519161646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,8192,0.024511999554104272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,6144,0.04284088810284933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,6144,0.01961599952644772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,5120,0.020672889219390023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,5120,0.04096622268358866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,4096,0.015038222074508667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,4096,0.019448000523779128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,3584,0.013395555317401886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,4096,0.0388008885913425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,3584,0.018766222728623282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,3072,0.012574222352769641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,2560,0.009071110851234859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,3072,0.018703111343913607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,3584,0.0377146667904324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,3072,0.03613244493802389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,2560,0.01773599949147966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,2048,0.008015111088752747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,65536,0.09157244364420573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,2560,0.0351217786471049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,2048,0.016386666231685214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,1536,0.006867555694447623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,2048,0.03334044416745504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,1536,0.014371555712487964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,1024,0.013747555514176687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,1024,0.005489777773618698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,1536,0.030085331863827173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,1024,0.028012444575627644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,768,0.004396444393528832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,768,0.013636444177892474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,512,0.003645333151022593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,768,0.027169777287377253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,512,0.013411555853154926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,256,0.0032764443506797156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,512,0.02678577767478095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,65536,0.15135378307766384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,256,0.013037333057986366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,128,0.003059555465976397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,128,0.025715554753939312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,256,0.025871111287011042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,128,0.012950222525331708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,64,0.002875555513633622
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,32,0.002954666606254048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,64,0.012724444270133972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,32,0.013254221942689685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,65536,0.14690044191148546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,16384,0.03247733248604669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,16384,0.06103110975689358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,12288,0.027497778336207073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,12288,0.053616000546349414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,16384,0.034640888373057045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,10240,0.02479022244612376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,10240,0.05063199996948242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,12288,0.027719999353090923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,8192,0.02272266646226247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,10240,0.02426044477356805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,8192,0.044495999813079834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,8192,0.02047911120785607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,7168,0.02165422174665663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,6144,0.016689777374267578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,7168,0.04372355673048231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,6144,0.020784000555674236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,5120,0.01475911173555586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,7168,0.018580444984965853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,6144,0.042259554068247475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,5120,0.019715555840068393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,65536,0.08191021945741442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,4096,0.012833777401182385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,5120,0.040932443406846784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,4096,0.01887022289964888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,65536,0.14028888278537327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,3584,0.012075555821259817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,3584,0.018530666828155518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,3072,0.011211555865075854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,4096,0.03894933395915561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,3584,0.03745955559942458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,3072,0.017807111144065857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,3072,0.035818666219711304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,2560,0.00850844466023975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,2560,0.016550223032633465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,2048,0.007173333730962541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,1536,0.006500444478458828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,2560,0.034830222527186074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,2048,0.014857777290874057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,1024,0.004433777597215441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,1536,0.014147554834683737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,2048,0.033246222469541765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,768,0.003934222377008862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,1024,0.01370933320787218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,1536,0.029727111260096233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,768,0.01365066650840971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,512,0.003604444364706675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,1024,0.027656000521447923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,768,0.026926222774717543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,512,0.01331022216214074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,256,0.0032711111836963226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,512,0.026425777210129633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,256,0.013032000097963544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,128,0.0029973333908451926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,256,0.026360889275868733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,128,0.012560000022252401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,64,0.002883555574549569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,32,0.0029155556112527847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,128,0.02563200063175625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,64,0.012640888492266336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,32,0.013219555219014486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,65536,0.12169155809614395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,12288,0.02548800077703264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,16384,0.029926220575968426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,12288,0.05304533243179321
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,10240,0.023177777727444965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,16384,0.05857511361440023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,12288,0.022799111074871484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,16384,0.02867377797762553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,10240,0.048839999569786914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,8192,0.021477333373493616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,10240,0.020263999700546265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,7168,0.015656888484954834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,8192,0.04376888937420315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,7168,0.02054933375782437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,6144,0.014009777042600842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,7168,0.04346844553947449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,8192,0.016954667038387723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,6144,0.019903111788961623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,5120,0.012795555922720166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,6144,0.04073422153790792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,5120,0.018975999620225694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,65536,0.07386489046944512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,4096,0.011164444188276926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,4096,0.01828799976242913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,5120,0.03979288869433933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,3584,0.010653333531485664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,4096,0.03783822059631348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,3584,0.017131555411550734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,3072,0.010088889135254754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,3584,0.03711466656790839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,3072,0.016504888733228047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,2560,0.007834666305118138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,3072,0.03548266821437412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,2560,0.016169778174824186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,2048,0.006798222247097228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,2560,0.03432000014517043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,2048,0.014522666732470194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,1536,0.005694222119119432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,2048,0.031700445546044245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,1536,0.014243554737832812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,1024,0.004316444612211651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,1536,0.029397332006030615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,65536,0.14856266975402832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,1024,0.013588444226317935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,768,0.0038888889054457345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,1024,0.027757333384619817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,768,0.013435555828942193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,512,0.003547555456558863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,768,0.02682133350107405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,512,0.013188444077968597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,256,0.0032177778581778207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,512,0.026703110999531213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,256,0.012824888858530255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,128,0.012612444659074148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,128,0.003003555453485913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,256,0.025995555851194594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,64,0.0028826666788922418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,32,0.0028533333291610083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,64,0.01254755589697096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,128,0.025659556190172832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,32,0.013173333472675748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,16384,0.02752711044417487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,16384,0.05687466594907972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,12288,0.02350133326318529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,65536,0.0980808867348565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,12288,0.01891288823551602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,16384,0.023450666003757056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,12288,0.05033333433998955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,10240,0.01685599982738495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,10240,0.022190221481853064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,8192,0.014340443743599786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,8192,0.020086222224765353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,10240,0.047968000173568726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,7168,0.013447999954223633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,8192,0.04392355680465698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,7168,0.019506666395399306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,6144,0.012324444121784635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,6144,0.019319999549123976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,7168,0.04285866684383816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,5120,0.01142133358452055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,5120,0.018405333161354065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,6144,0.040652443965276085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,65536,0.06748088863160875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,4096,0.010018666585286459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,65536,0.13721599843766955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,5120,0.03926577832963731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,4096,0.016188444362746347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,3584,0.009546666509575313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,4096,0.03772355450524224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,3584,0.015806222955385845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,3072,0.009220444493823582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,3072,0.01556799974706438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,2560,0.00757422215408749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,3584,0.03619199991226196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,2560,0.014919999572965833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,3072,0.033313777711656355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,2048,0.0063164445261160536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,2560,0.03198755449718899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,2048,0.014459555347760519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,1536,0.005014222115278244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,1536,0.02905511193805271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,2048,0.030983110268910725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,1536,0.014053333136770459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,1024,0.0041582224269707995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,1024,0.013748444616794586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,768,0.0038648889296584656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,1024,0.027630221512582567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,768,0.01346666696998808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,512,0.003526222374704149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,768,0.027383110589451257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,512,0.013223999904261695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,256,0.0031991110493739447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,512,0.026550221774313185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,256,0.012929778132173749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,128,0.002983111060327954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,256,0.0258240004380544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,128,0.0129120002190272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,64,0.0028435554769304064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,65536,0.07538666990068224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,32,0.0028213332924577924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,128,0.02572444412443373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,64,0.012597333225939008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,32,0.01313066648112403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,16384,0.024971556332376268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,12288,0.01440622243616316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,12288,0.02200800014866723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,16384,0.05202310946252611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,10240,0.013342222405804528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,12288,0.04749955402480232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,10240,0.02091022166940901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,16384,0.017492444978819955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,8192,0.011573333707120685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,10240,0.04511555698182848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,8192,0.01973511113060845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,7168,0.011078221930397881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,8192,0.04051911168628269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,7168,0.018965333700180054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,6144,0.0098213329911232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,7168,0.041025777657826744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,6144,0.01815111107296414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,5120,0.009508444203270806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,5120,0.01714666684468587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,6144,0.03977866636382209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,65536,0.05550133188565572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,4096,0.008938666847017076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,5120,0.03867466582192315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,4096,0.015807999504937064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,3584,0.008935111264387766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,4096,0.034550223085615374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,3584,0.015557333827018738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,3072,0.007970666719807519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,65536,0.11910133891635472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,3584,0.033591111501057945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,3072,0.015400000744395785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,2560,0.006624889042642381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,3072,0.03216444452603658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,2560,0.014901333385043673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,2048,0.005879999862776862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,2560,0.030598223209381104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,2048,0.014785778191354541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,1536,0.014162666267818876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,1536,0.005141333159473208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,2048,0.03003822101487054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,1536,0.02868355645073785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,1024,0.0041804446114434134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,768,0.0037919998996787598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,65536,0.050811555650499135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,1024,0.013893332746293811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,512,0.003677333394686381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,512,0.013188444077968597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,1024,0.027169777287377253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,768,0.013551111022631327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,768,0.02671555512481266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,256,0.003237333355678452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,256,0.012843555874294706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,512,0.02611733310752445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,128,0.002983111060327954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,256,0.025648888614442613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,128,0.012713777522246042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,64,0.0028533333291610083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,128,0.025711110896534387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,32,0.0028293333533737394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,64,0.012537777423858643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,32,0.013014222184816996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,16384,0.015091554986106025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,16384,0.02365511159102122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,12288,0.012523555921183692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,12288,0.02127200033929613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,16384,0.04700444473160637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,10240,0.011612444288200803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,10240,0.020269332660569083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,12288,0.04548799991607666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,8192,0.010306666294733683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,10240,0.04203377829657661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,8192,0.018466666340827942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,7168,0.009938666390048133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,7168,0.01851022243499756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,8192,0.03840000099605984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,6144,0.00904177791542477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,7168,0.038753777742385864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,6144,0.016785777277416654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,65536,0.05359199974271986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,5120,0.008912000391218398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,6144,0.03570577833387587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,5120,0.016516443755891588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,4096,0.007660444411966536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,65536,0.10783644517262776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,5120,0.034303999609417386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,4096,0.015815110670195687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,3584,0.008314666648705801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,65536,0.04078133238686456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,4096,0.033033778270085655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,3584,0.01568177839120229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,3072,0.0074177781740824384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,3584,0.03216977914174398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,2048,0.005532444351249271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,3072,0.015243555108706156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,2560,0.00639111093348927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,3072,0.030829333596759375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,2560,0.014764444695578681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,2560,0.030054221550623577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,1536,0.004892444445027245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,2048,0.014661333627170987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,2048,0.029861334297392104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,1536,0.014137778017255994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,1024,0.0041893331540955436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,1536,0.028394665982988145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,1024,0.013625777430004544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,768,0.0038613333470291565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,1024,0.027141332626342773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,768,0.013638222383128272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,512,0.003491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,768,0.02682400080892775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,512,0.013249778085284762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,256,0.0031831111345026228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,256,0.01295466638273663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,512,0.026105778084860906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,128,0.0029884444342719186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,256,0.025735111700163946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,128,0.01275644451379776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,64,0.0028168888141711554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,128,0.025634666283925373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,32,0.00281866660548581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,64,0.012635555532243518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,32,0.01348355578051673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,16384,0.011539555258221097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,16384,0.023387556274731953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,12288,0.010310221877362993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,16384,0.046042667494879834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,12288,0.020990222692489624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,10240,0.009904888768990835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,12288,0.043256001340018384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,10240,0.01905866629547543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,8192,0.009242666264375051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,10240,0.039839999543295965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,65536,0.05151999990145365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,8192,0.017786666750907898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,65536,0.10159022278255886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,7168,0.008577778107590145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,7168,0.017288888494173687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,8192,0.03577422102292379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,6144,0.008049777812427944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,7168,0.03580177823702494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,6144,0.016538666354285348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,5120,0.007801777786678738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,65536,0.02972444560792711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,6144,0.034833778937657676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,5120,0.016169778174824186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,4096,0.007055111229419708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,5120,0.03348177671432495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,4096,0.015671110815472074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,3584,0.007310222420427535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,4096,0.032582223415374756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,3584,0.015583111180199517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,3072,0.0069288888739215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,3584,0.031535999642478094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,3072,0.01513066722287072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,2560,0.006127999888526068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,3072,0.030538668235143025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,1536,0.0047893333766195504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,2560,0.014827556080288358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,2048,0.005490666462315454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,2560,0.03046222196684943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,2048,0.014384888940387301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,2048,0.029247111744350855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,1536,0.013988445202509562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,1024,0.0041742221348815495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,1024,0.01370488852262497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,1536,0.02842666705449422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,768,0.003826666623353958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,1024,0.02735911144150628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,768,0.01334755536582735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,512,0.0035075553589397003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,768,0.026328888204362657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,256,0.013020444247457715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,512,0.013304000099500021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,256,0.0031644445326593188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,512,0.026133333643277485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,128,0.002961777771512667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,256,0.025610667135980394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,128,0.012691555751694573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,64,0.0028417776856157514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,128,0.025444444682863023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,32,0.0028373334142896864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,64,0.012619555824332766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,32,0.013008889224794177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,65536,0.018435556027624343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,16384,0.008951110972298516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,65536,0.04992977778116862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,16384,0.022416000564893086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,65536,0.1006773312886556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,12288,0.008252444366614023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,16384,0.044053332673178784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,12288,0.019606222709019978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,10240,0.007990222010347579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,10240,0.018564444449212816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,12288,0.0406257775094774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,8192,0.007146666447321574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,10240,0.03829511006673177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,8192,0.017846221725145977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,7168,0.006837333242098491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,8192,0.035263111193974815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,7168,0.01704355577627818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,6144,0.006772444479995304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,7168,0.03551022211710612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,6144,0.016671111186345417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,5120,0.007159999675220913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,6144,0.03446133269204034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,5120,0.016167110866970487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,4096,0.00675022229552269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,5120,0.03329688972897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,4096,0.01568800045384301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,3584,0.00720888872941335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,4096,0.03222399950027466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,3584,0.015507555670208402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,3072,0.006847111301289664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,3584,0.03162044286727905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,2560,0.014656000667148165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,3072,0.015019555886586508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,2560,0.006129777679840724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,3072,0.03050133254792955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,2048,0.005414222263627582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,2560,0.030099554194344416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,1536,0.004818666726350784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,2048,0.014409777190950183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,1536,0.02792177928818597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,2048,0.029162665208180744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,1536,0.01408088869518704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,1024,0.004119111018048393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,1024,0.01368622233470281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,768,0.0037475555307335327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,768,0.013403555585278405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,1024,0.02715733316209581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,512,0.0034719999465677473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,768,0.02628177735540602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,512,0.013192888763215808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,256,0.003125333330697484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,512,0.026161778304311965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,256,0.01292888902955585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,128,0.002942222274012036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,256,0.02623733381430308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,128,0.012527111503813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,64,0.0027822222974565295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,32,0.013070222404268054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,32,0.0027742222365405825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,128,0.02555022140343984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,64,0.012612444659074148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,65536,0.012820444173283048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,16384,0.006917333437336816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,65536,0.04900177650981479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,16384,0.021432888176706102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,12288,0.007002666592597961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,65536,0.09965333011415269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,16384,0.04334577918052673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,12288,0.019302222463819716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,10240,0.007088888850477006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,12288,0.040417777167426214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,10240,0.01831822262869941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,8192,0.00692622239391009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,10240,0.038303110334608294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,8192,0.017362667454613578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,7168,0.006736889067623351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,8192,0.035608887672424316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,7168,0.03552177879545424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,7168,0.016937777400016785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,6144,0.006525333142942852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,5120,0.006878222028414409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,6144,0.01646133263905843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,4096,0.006488889041874144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,5120,0.015841777125994366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,6144,0.03462400039037069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,4096,0.01553422212600708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,5120,0.03356444504525926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,3584,0.0069475554757648045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,4096,0.0321742229991489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,3584,0.015159999330838522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,3072,0.006608888920810487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,3584,0.03148622314135233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,3072,0.014892444014549255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,2560,0.005943111247486538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,3072,0.03038133184115092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,2560,0.014560888210932413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,2048,0.0053102220926019884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,2560,0.029927998781204224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,2048,0.014209777116775513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,1536,0.004661333229806689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,1024,0.013616888887352414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,2048,0.02885422110557556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,1536,0.013972444666756524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,1024,0.004049777984619141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,1536,0.027882665395736694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,768,0.0037235555549462638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,1024,0.026545777916908264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,768,0.013339555925793119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,512,0.003444444388151169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,768,0.02648799949222141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,512,0.013365333278973898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,256,0.0030817778574095834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,512,0.025796444879637823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,256,0.012744888663291931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,128,0.0028524444335036804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,256,0.025461332665549383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,128,0.012681778106424542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,64,0.0028106667515304354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,32,0.002720000015364753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,64,0.012705778082211813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,32,0.013033777475357056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,12288,0.006640888750553131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,65536,0.01012711144155926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,16384,0.006768888897365994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,65536,0.048730668094423085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,16384,0.02124799953566657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,10240,0.006774222271309958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,12288,0.01907999979125129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,8192,0.006629333313968446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,10240,0.017990221579869587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,7168,0.006535111202134027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,8192,0.017156443662113614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,7168,0.01661688917213016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,6144,0.006441777778996362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,5120,0.006767999794748094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,6144,0.016043555405404832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,4096,0.006416888700591193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,5120,0.015621332658661736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,3072,0.006474666711356904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,3584,0.006871999965773688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,4096,0.015062222878138224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,3584,0.014954666296641031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,2560,0.0057884446448749965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,3072,0.0147724449634552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,2048,0.00519466648499171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,2560,0.014335999886194864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,1536,0.004585777719815572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,2048,0.014077333940400017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,1024,0.003951999876234266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,1536,0.013735111388895245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,768,0.003608889049953885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,1024,0.013421333498424955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,512,0.0033413333197434745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,768,0.013251555462678274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,256,0.003030222323205736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,512,0.01312622262371911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,128,0.002883555574549569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,256,0.012786666552225748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,64,0.0027520000520679685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,128,0.012444444828563266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,32,0.0027457777824666765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,64,0.012425777812798819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,32,0.012979555461141797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,65536,0.00962488849957784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,16384,0.006878222028414409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,65536,0.04754222101635403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,12288,0.006784889019197888
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,16384,0.021274665991465252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,10240,0.006667555620272954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,12288,0.019215111931165058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,8192,0.006648889018429651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,10240,0.018138666947682697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,7168,0.0064995553758409284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,8192,0.017243555850452848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,6144,0.006364444477690591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,7168,0.01681866745154063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,5120,0.006704000135262807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,6144,0.016359999775886536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,4096,0.006360000206364526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,5120,0.015793777174419828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,3584,0.006775110960006714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,4096,0.015287111202875773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,3072,0.006467555546098285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,3584,0.014884443746672736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,2560,0.005767999837795894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,3072,0.014816888504558139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,2048,0.005138666679461797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,2560,0.014422222971916199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,1536,0.004598222259018156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,2048,0.014007111390431723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,1024,0.003947555604908201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,512,0.003350222276316749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,1536,0.0136835558546914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,768,0.003630222131808599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,1024,0.01346666696998808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,768,0.013235555754767524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,256,0.0030213333666324615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,512,0.013072888884279462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,128,0.002855111120475663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,32,0.002708444371819496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,256,0.013059555656380124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,64,0.0027884443600972495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,128,0.012515555653307172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,64,0.012511110968059964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,32,0.012960889273219638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,12288,0.2802986568874783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,16384,0.3677528964148627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,12288,0.3492862118615045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,16384,0.41857245233323837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,10240,0.23595200644599068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,10240,0.47349956300523544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,10240,0.2988426685333252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,8192,0.19197332859039307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,128,0.0255724440018336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,12288,0.5709217919243706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,8192,0.24427466922336152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,7168,0.17089245054456922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,8192,0.3789057731628418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,6144,0.1714951064851549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,7168,0.20688177479637992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,16384,0.7454293568929037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,6144,0.1894071102142334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,7168,0.330168883005778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,4096,0.11911288897196452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,5120,0.14466222127278647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,6144,0.2852693398793538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,5120,0.16490044858720568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,5120,0.23765688472323945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,4096,0.19226933850182426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,4096,0.12654133637746176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,3584,0.09448444181018406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,3072,0.09355377488666111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,3584,0.11951288912031387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,3072,0.10092710786395603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,3584,0.1678160031636556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,2560,0.07301600111855401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,2560,0.12166044447157119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,2048,0.061724444230397545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,2560,0.09115911192364162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,2048,0.08251555760701497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,3072,0.14467822180853948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,2048,0.09723822275797527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,1536,0.05138044555981954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,1024,0.04148088892300924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,1536,0.06528888808356391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,768,0.037929776642057635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,1024,0.050888889365726046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,1024,0.05466577741834852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,512,0.033835556772020124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,768,0.04888888862397936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,512,0.027531555957264368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,256,0.02885599931081136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,128,0.00923822240697013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,512,0.04193066557248434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,256,0.03589866558710734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,128,0.02676800058947669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,768,0.0458204448223114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,256,0.015472888946533203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,1536,0.08703200022379558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,64,0.006507555643717448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,128,0.03252799974547492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,32,0.004566222015354368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,64,0.024324445260895625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,32,0.023190221852726404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,12288,0.08361777994367812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,16384,0.10563821925057305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,16384,0.1290186643600464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,12288,0.14093244075775146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,10240,0.07225333319769965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,12288,0.10643911361694336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,10240,0.11664266056484646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,10240,0.09894311428070068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,16384,0.184843553437127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,8192,0.06110666857825386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,8192,0.07967288626564874
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,8192,0.09458400143517388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,7168,0.05563466416464912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,7168,0.08320533567004733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,7168,0.07293244202931722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,6144,0.05015733175807529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,6144,0.06720266739527385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,6144,0.07262488868501452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,5120,0.0448035564687517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,5120,0.060639109876420766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,5120,0.060455110337999135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,4096,0.03913333349757724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,4096,0.05508888761202494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,3584,0.03958311014705234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,3584,0.05223200056287977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,4096,0.04936710993448893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,3584,0.04358222087224325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,3072,0.03632622294955783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,3072,0.049136890305413135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,2560,0.03306577934159173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,2560,0.0458497769302792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,3072,0.037887109650505915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,2560,0.03209866748915778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,2048,0.027781334188249376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,1536,0.024519999821980793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,2048,0.026276444395383198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,2048,0.04217244519127739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,1536,0.03731377919514974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,1024,0.021570665968788996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,65536,0.3718453248341878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,768,0.011651555697123209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,1536,0.020776889390415616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,768,0.020042666130595736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,1024,0.03442755672666762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,512,0.008757332960764566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,512,0.018644443816608854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,768,0.033619556162092425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,256,0.006121777825885349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,1024,0.014661333627170987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,512,0.03135644396146139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,256,0.014545778433481852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,128,0.003556444413132138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,256,0.026791110634803772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,128,0.014308444327778287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,64,0.003302222324742211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,128,0.026124444272783067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,32,0.0032675556010670135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,64,0.01497866710027059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,32,0.014545778433481852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,65536,0.42628712124294704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,16384,0.08465688758426243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,16384,0.11329155498080784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,12288,0.06765421893861559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,16384,0.1458666721979777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,12288,0.09195378091600205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,12288,0.11195377508799235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,10240,0.05919910801781548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,10240,0.08013955752054851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,65536,0.29006756676567924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,8192,0.05055555701255798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,10240,0.0885537796550327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,8192,0.07022044393751356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,8192,0.08576266633139716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,7168,0.046132445335388184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,7168,0.0652302238676283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,7168,0.06182222233878242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,6144,0.04224444429079691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,6144,0.054679112301932446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,5120,0.041647109720442034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,6144,0.06030133035447863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,4096,0.033879998657438494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,5120,0.055155555407206215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,5120,0.04555288950602213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,4096,0.05015644431114197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,4096,0.037348445918824934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,3584,0.03173866536882188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,65536,0.3609813319312201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,3584,0.047600001096725464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,3584,0.03586133321126302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,3072,0.02947910957866245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,2560,0.027089777919981215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,2048,0.0244835548930698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,2560,0.04253333475854662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,3072,0.02929511004024082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,3072,0.045055111249287925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,2048,0.03790222273932563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,1536,0.022651554809676275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,2048,0.021664889322386846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,1536,0.03533777925703261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,2560,0.025591111845440332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,1024,0.01182755579551061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,65536,0.7501760058932834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,768,0.009830222361617619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,1536,0.01607466737429301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,1024,0.02055199940999349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,1024,0.032919999625947736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,768,0.01939733326435089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,512,0.007640000018808577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,768,0.03193155593342251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,512,0.01734133395883772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,256,0.004423999951945411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,512,0.030419554975297716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,256,0.014389332797792224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,128,0.003532444437344869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,256,0.026762665973769292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,128,0.014160000615649752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,64,0.0032844444115956626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,128,0.02609155575434367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,64,0.014925332532988654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,32,0.0032871110985676446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,32,0.014081777797804939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,16384,0.07431111070844862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,16384,0.09636000129911636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,65536,0.5622853173149956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,12288,0.05997155772315132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,12288,0.09347733524110581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,12288,0.08231821987364027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,16384,0.11841332912445068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,10240,0.05237244566281637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,10240,0.07630933655632867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,10240,0.07096978028615315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,8192,0.04526933365397983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,8192,0.0625564455986023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,8192,0.060956445005204946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,7168,0.04211644331614176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,7168,0.05380622214741177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,7168,0.058992001745435924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,6144,0.03851022322972616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,6144,0.047035555044809975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,6144,0.055142223834991455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,5120,0.03591022226545546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,5120,0.04526577724350823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,5120,0.05082044336530897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,4096,0.032200889454947576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,4096,0.03156355685657925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,3584,0.02961155441072252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,3072,0.027475555737813313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,4096,0.04667199982537163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,3584,0.02867822183503045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,3584,0.04468889037768046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,3072,0.042598221037122935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,65536,0.24344176716274687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,2560,0.02539199921819899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,2560,0.03929422299067179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,2048,0.023412444525294836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,1536,0.013793778088357715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,3072,0.025891555680169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,2048,0.03647111190689935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,1536,0.02163022259871165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,1024,0.010512000156773461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,1536,0.03445422318246629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,2048,0.017519111434618633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,2560,0.021351999706692163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,1024,0.019740444090631273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,768,0.008683555656009251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,1024,0.03256888853179084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,768,0.01850488947497474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,768,0.031684445010291204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,512,0.015486222174432544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,512,0.006809777683681912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,256,0.003810666501522064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,512,0.027468444572554693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,256,0.014308444327778287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,128,0.0033093332830402586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,256,0.027007111244731482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,128,0.01405066748460134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,64,0.0032337777730491427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,128,0.026036444637510512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,64,0.014715555641386243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,32,0.003120888852410846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,32,0.013997332917319404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,65536,0.31614934073554146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,16384,0.06155111392339071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,16384,0.08462133010228474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,12288,0.050627556112077504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,12288,0.0708035561773512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,16384,0.09272533655166626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,12288,0.06971377796596952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,10240,0.04455999864472283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,10240,0.05910578038957384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,10240,0.06386044290330675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,8192,0.03892088929812113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,8192,0.05650133556789822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,8192,0.04797244403097364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,7168,0.035990221632851496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,7168,0.0532604455947876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,7168,0.042600890000661217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,6144,0.03310577736960517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,6144,0.037301331758499146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,65536,0.25575023227267796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,65536,0.19659466213650176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,6144,0.05006400081846449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,5120,0.030200888713200886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,65536,0.47643375396728516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,4096,0.04304711023966471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,3584,0.027274666561020747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,4096,0.02698933415942722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,3584,0.04061333338419596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,5120,0.04676533407635159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,5120,0.031643556223975286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,4096,0.025803556044896443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,3072,0.022855111294322546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,3584,0.023015111684799194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,3072,0.039204445150163435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,2560,0.02316177719169193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,2560,0.03884977764553494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,2048,0.019928889142142404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,1536,0.011639110743999481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,2048,0.03503111004829407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,1536,0.018562666243977018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,1024,0.008752000000741746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,2560,0.017479111750920612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,3072,0.020392000675201416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,1536,0.032997333341174655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,2048,0.014447111222479077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,1024,0.016992888516849942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,768,0.007459555235173967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,1024,0.03167111012670729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,768,0.014108444253603617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,512,0.00599911105301645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,768,0.030685332086351182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,512,0.013601777454217276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,256,0.003624000069167879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,512,0.02678844498263465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,256,0.012850667039553324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,128,0.0032506665835777917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,256,0.02629866699377696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,128,0.01273333364062839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,64,0.003134222287270758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,128,0.02590222160021464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,32,0.003015111097031169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,64,0.01348355578051673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,32,0.012864000267452665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,16384,0.05875377522574531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,16384,0.07672888702816434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,12288,0.06594400273429023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,12288,0.04667288727230496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,12288,0.06738133562935723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,16384,0.08487911356820001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,10240,0.04128977656364441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,10240,0.055882665846082896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,10240,0.06060889032151964
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,8192,0.03626577721701728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,8192,0.05525066786342197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,8192,0.04583911101023356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,65536,0.37444265683492023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,7168,0.033583111233181424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,7168,0.0410791105694241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,7168,0.05198311143451267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,6144,0.031160000297758315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,6144,0.04936088787184822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,5120,0.031192888816197712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,6144,0.03654222355948554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,5120,0.028492444091373022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,65536,0.18230844868554008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,4096,0.024640889631377325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,5120,0.047707554366853505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,4096,0.023335999912685815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,4096,0.0414613319767846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,3584,0.023176888624827068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,3584,0.03923733366860284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,3072,0.021935110290845234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,3584,0.02095111045572493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,65536,0.23319289419386122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,3072,0.03748622205522325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,2560,0.020618667205174763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,3072,0.01867733399073283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,2048,0.013249778085284762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,2560,0.015996444556448195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,2560,0.03663999835650126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,2048,0.019191111127535503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,2048,0.03457066747877333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,1536,0.010832889212502373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,1536,0.018950222267044913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,1024,0.008208889100286696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,1536,0.03272355596224467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,1024,0.01572088897228241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,768,0.006791111081838608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,1024,0.03138755427466498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,768,0.013873777455753751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,768,0.03067022230890062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,512,0.005573333137565189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,512,0.013487999637921652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,512,0.02683199942111969
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,256,0.0034204444123639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,256,0.012871999707486896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,256,0.026213333010673523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,128,0.0032293332947625052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,128,0.012757333616415659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,128,0.025807110799683466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,64,0.002992000016901228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,64,0.013321778012646569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,32,0.002995555599530538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,32,0.012723555167516073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,65536,0.32388533486260307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,16384,0.05259733398755392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,16384,0.07289244731267293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,12288,0.047228445609410606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,12288,0.0584826668103536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,12288,0.06368266873889498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,16384,0.07622133360968696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,10240,0.03869244456291199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,10240,0.058740443653530545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,10240,0.047080000241597496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,8192,0.0341360006067488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,8192,0.03913777735498216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,8192,0.052778667873806424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,7168,0.03201155530081855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,7168,0.03427377674314711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,7168,0.05059022373623318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,6144,0.029663112428453233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,6144,0.047409776184293956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,5120,0.026887110537952844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,6144,0.030085331863827173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,65536,0.1631422175301446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,5120,0.04448622133996752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,4096,0.02405244443151686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,4096,0.03999911083115472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,3584,0.022482666704389784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,5120,0.025555556019147236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,4096,0.021420444051424663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,3584,0.03868800070550706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,3584,0.01871200071440803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,3072,0.021192888418833416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,3072,0.016924444172117446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,3072,0.03707733419206407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,65536,0.20666221777598062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,2560,0.014730667074521383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,2560,0.019878221882714164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,2048,0.012542222109105853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,2560,0.0358897778722975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,2048,0.01871200071440803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,2048,0.03379822108480666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,1536,0.011071110765139261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,1536,0.017664000391960144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,1536,0.03237422307332357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,768,0.014736889137162102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,1024,0.007659555309348636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,1024,0.015023999743991427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,768,0.006506666541099548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,1024,0.03119644522666931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,512,0.004651555584536659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,768,0.028020444843504164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,512,0.013401777380042605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,128,0.012707555459605323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,256,0.0033324443631701996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,512,0.02676800058947669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,256,0.012789333032237159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,256,0.026221333278550044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,128,0.0031999999450312722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,64,0.0029884444342719186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,128,0.02591288917594486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,64,0.013351110948456658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,32,0.0029608888758553397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,32,0.012622222304344177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,65536,0.2843200100792779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,16384,0.045891556474897593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,16384,0.0690666635831197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,12288,0.04731377628114489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,16384,0.060232890976799854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,12288,0.038979556825425886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,12288,0.06175288889143202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,10240,0.03492177857293023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,8192,0.03137066629197862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,10240,0.05739466349283854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,8192,0.0317520002524058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,10240,0.04023377762900458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,8192,0.051891555388768516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,7168,0.02953688965903388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,7168,0.028174221515655518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,6144,0.02714577813943227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,65536,0.13844266202714708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,65536,0.18354666233062744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,7168,0.05000977714856466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,5120,0.024562666813532513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,6144,0.04565599891874525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,5120,0.042725334564844765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,4096,0.02221244408024682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,6144,0.028734223710166082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,4096,0.039487110243903265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,5120,0.021324444148275588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,4096,0.01767200065983666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,3584,0.021291555629836187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,3584,0.03801777627733018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,3072,0.014173333843549093
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,3584,0.016036444240146212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,3072,0.020303999384244282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,3072,0.036722666687435575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,2560,0.012248888611793518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,2560,0.019144000278578866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,2560,0.03552800085809495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,2048,0.010453333457310995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,2048,0.01817599932352702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,2048,0.033293333318498396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,1536,0.0085235552655326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,1536,0.017704889178276062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,1024,0.007044444481531779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,1536,0.031939556201299034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,1024,0.013993778162532382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,65536,0.2360346582200792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,768,0.006152888966931238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,1024,0.03049066662788391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,768,0.013616888887352414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,768,0.027758222487237718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,512,0.003834666477309333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,512,0.013316444224781461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,512,0.027095110880004034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,256,0.003306666596068276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,256,0.01274400038851632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,128,0.0031191110610961914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,256,0.026345777842733595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,128,0.012629333469602795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,128,0.025832888152864244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,64,0.0029351111087534162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,64,0.013458666702111563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,32,0.0029191111938820947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,32,0.012506666282812754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,16384,0.040398223532570735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,12288,0.034499555826187134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,16384,0.06577510966195001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,12288,0.04012088974316915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,16384,0.05095644460784065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,12288,0.05935821930567423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,10240,0.03109955456521776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,10240,0.054663111766179405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,10240,0.03412444392840067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,8192,0.028322666883468628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,8192,0.028351111544503108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,7168,0.026595556073718604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,8192,0.05008977651596069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,7168,0.04633333285649618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,6144,0.024153777294688757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,7168,0.025571554899215698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,65536,0.11452178160349528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,6144,0.04404800136884054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,6144,0.0229120006163915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,65536,0.1740960015190972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,5120,0.022224000758594934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,5120,0.040568000740475126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,4096,0.020689777202076383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,3584,0.01329066687160068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,5120,0.020249777370029025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,4096,0.03848888807826572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,3584,0.019951999187469482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,3584,0.0370417767100864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,3072,0.01198755535814497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,3072,0.01978755493958791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,4096,0.016925333274735343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,2560,0.010471999645233154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,3072,0.035747554567125105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,2560,0.01829866733815935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,2560,0.03490488727887472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,2048,0.009026666482289633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,2048,0.01717688971095615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,1536,0.007655999726719326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,2048,0.03280177712440491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,1536,0.016043555405404832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,1536,0.031529777579837374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,1024,0.006256888724035687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,1024,0.013722666435771518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,1024,0.028543998797734577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,768,0.004875555634498596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,768,0.01350933313369751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,768,0.027287110686302185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,512,0.0036959999965296853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,512,0.013251555462678274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,65536,0.1878195603688558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,512,0.026504889130592346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,256,0.003305777700410949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,256,0.012815111213260226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,256,0.0262106673585044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,128,0.0031191110610961914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,128,0.01256622208489312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,64,0.002879111096262932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,128,0.02569777766863505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,64,0.013191111385822296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,32,0.0028986665937635633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,32,0.012438221938080259
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,16384,0.03795466820398966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,12288,0.03257066673702664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,16384,0.06472444534301758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,12288,0.03571733170085483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,16384,0.045808000697029956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,10240,0.029662221670150757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,12288,0.057725330193837486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,8192,0.026757333013746474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,8192,0.04867911007669237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,10240,0.05454755491680569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,8192,0.02572888963752323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,10240,0.030962665875752766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,7168,0.024671110841963027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,7168,0.045688887437184654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,6144,0.02318488889270359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,65536,0.1018844445546468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,7168,0.023426666855812073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,6144,0.04287733303176033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,5120,0.02149244480662876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,4096,0.01331377774477005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,4096,0.020436444216304354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,5120,0.04053333401679993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,4096,0.03853066762288412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,6144,0.02084000077512529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,3584,0.012164444559150271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,3584,0.019688000281651814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,3072,0.011016888750924004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,5120,0.018386666973431904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,3584,0.036736889017952815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,3072,0.018616000811258953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,2560,0.00981155534585317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,3072,0.03574400146802267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,2560,0.01791733337773217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,2048,0.008270222279760573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,2560,0.03455377618471781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,2048,0.01692800058258904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,2048,0.03259377678235372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,1024,0.006066666709052191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,1536,0.007057777709431118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,1536,0.014249778456158109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,1536,0.03128711051411099
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,1024,0.013696889082590738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,768,0.004392888810899523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,65536,0.16272889243231878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,1024,0.02772888872358534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,768,0.013378666506873237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,512,0.0036311112344264984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,768,0.027168888184759352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,512,0.013153778182135688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,256,0.026047110557556152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,256,0.003338666632771492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,512,0.02641244398223029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,256,0.012837332983811697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,128,0.003048000029391713
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,128,0.025595555702845257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,64,0.0028364445186323593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,128,0.012585777375433179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,32,0.0028648889727062653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,64,0.013239999612172445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,32,0.0123831108212471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,16384,0.03414933217896356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,16384,0.06195911433961657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,12288,0.03131377696990967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,16384,0.04065955678621928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,12288,0.02962133288383484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,65536,0.17536621623569063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,10240,0.027033777700530157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,12288,0.057244446542527944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,10240,0.05375822385152181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,8192,0.024102222588327196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,8192,0.04699555701679654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,7168,0.022913777165942725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,65536,0.0884808897972107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,10240,0.027217777238951787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,7168,0.0455715556939443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,6144,0.02165777815712823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,8192,0.023196443915367126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,7168,0.020990222692489624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,6144,0.04238666759596931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,5120,0.016474665866957772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,6144,0.01870844430393643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,4096,0.014715555641386243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,5120,0.040456000301573015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,5120,0.02027466727627648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,3584,0.01370133293999566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,4096,0.019317333896954853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,4096,0.037866665257347956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,65536,0.1515422264734904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,3584,0.0188071115149392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,3072,0.012431999875439538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,3584,0.03677689035733541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,3072,0.01814933286772834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,2560,0.00905600024594201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,3072,0.03568089008331299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,2560,0.017270222306251526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,2048,0.00797333319981893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,2560,0.03452533483505249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,1536,0.006741333338949416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,2048,0.014890667464998035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,2048,0.03206311000718011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,1536,0.01420088940196567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,1024,0.005280888742870755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,1536,0.029751112063725788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,1024,0.013694222602579327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,768,0.004327110946178436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,1024,0.02829688787460327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,512,0.0037253333462609183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,768,0.01351111133893331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,512,0.013213333156373767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,768,0.02732622292306688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,256,0.012930666406949362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,256,0.0032968889507982465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,512,0.02650044361750285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,128,0.003080000066094928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,256,0.025976000560654536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,128,0.01255555533700519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,64,0.0028568889117903183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,128,0.026064000195927087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,32,0.0029066666546795103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,64,0.013086222112178802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,32,0.01240711079703437
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,65536,0.14512711101108128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,16384,0.03116977877087063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,12288,0.027056889401541814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,12288,0.027520888381534155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,16384,0.06072355641259087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,10240,0.025610667135980394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,12288,0.05559111303753323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,10240,0.052348444859186806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,16384,0.0343413319852617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,8192,0.022555554906527202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,8192,0.047264890538321606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,7168,0.021597334080272253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,10240,0.023761777414215937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,8192,0.0203128887547387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,6144,0.01648177703221639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,7168,0.04490844408671061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,6144,0.020469332734743755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,5120,0.01458577811717987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,7168,0.01829866733815935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,6144,0.04168533285458883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,5120,0.019451555278566148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,65536,0.07734844419691297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,4096,0.012732444538010491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,5120,0.040087110466427274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,4096,0.01883288886811998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,3584,0.011969777444998423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,4096,0.03706577751371596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,3584,0.018152889278199937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,3072,0.011037333144081963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,3584,0.03668177790111966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,3072,0.035197334157096014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,3072,0.017643555998802185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,2560,0.008198222352398766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,2560,0.015929778416951496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,2048,0.007224889265166388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,2048,0.015384889311260648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,2560,0.03434222274356418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,1536,0.006539555473460092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,2048,0.02963199880388048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,1536,0.014087110757827759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,1024,0.004975111120276981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,1536,0.028868443436092798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,1024,0.01368622233470281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,768,0.0042071110672420925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,1024,0.02776177724202474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,768,0.01349244432316886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,512,0.003576000117593341
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,768,0.027121777335802715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,512,0.013129777378506131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,256,0.003248888792263137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,65536,0.13579466607835558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,512,0.026387555731667414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,256,0.012631110846996307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,128,0.003034666594531801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,256,0.026084444589085046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,128,0.012503110700183444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,64,0.00286666676402092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,128,0.025613332788149517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,64,0.013199999928474426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,32,0.002895111011134254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,32,0.01236355553070704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,16384,0.028966221544477675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,16384,0.059671110577053495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,12288,0.02494311167134179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,16384,0.028000000450346205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,65536,0.12103911240895589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,12288,0.053379555543263756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,10240,0.022848000129063923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,12288,0.022826666633288067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,10240,0.04997333221965366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,8192,0.02131822208563487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,10240,0.02000177734427982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,8192,0.046001778708563916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,7168,0.015305777390797934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,7168,0.020534222324689228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,6144,0.013814222481515674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,8192,0.016564443707466125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,7168,0.04439111219512092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,65536,0.06704711251788668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,6144,0.01975822283162011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,65536,0.14793777465820312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,6144,0.040933334165149264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,5120,0.012702222499582501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,5120,0.01903288894229465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,4096,0.01126399967405531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,4096,0.017462222112549674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,5120,0.038831111457612776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,3584,0.010563555690977307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,4096,0.03713599840799967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,3584,0.01756000022093455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,3072,0.009951111343171861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,3072,0.015951999359660678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,2560,0.00775999989774492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,3584,0.035535110367669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,3072,0.03532177872127957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,2560,0.01510933372709486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,1536,0.00599911105301645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,2048,0.006874666859706243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,2560,0.033815999825795494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,2048,0.014564444621404013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,2048,0.029234667619069416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,1536,0.014039999908871122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,1024,0.004401777767472797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,1536,0.028522666957643297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,1024,0.013556444810496436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,768,0.003843555433882607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,1024,0.027644443843099806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,768,0.027502222193611994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,768,0.013297778036859302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,512,0.003506666670242945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,512,0.013192888763215808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,256,0.0031822222388452957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,512,0.026214222113291424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,65536,0.09768799940745036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,256,0.012771555946932899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,128,0.0029200000895394217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,256,0.026364444030655756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,128,0.012498666842778524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,64,0.002814222127199173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,32,0.002809777855873108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,128,0.0254711111386617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,64,0.013136000268989138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,32,0.012320888539155325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,16384,0.026783999469545152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,12288,0.023494222097926672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,16384,0.057180444399515785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,10240,0.016692444682121277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,12288,0.052285333474477134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,10240,0.021912000245518152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,12288,0.018618666463428073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,16384,0.023103111320071753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,8192,0.014190221826235453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,10240,0.049152000082863696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,7168,0.013357333011097379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,8192,0.020444444484180875
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,7168,0.019698666201697457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,8192,0.04414666692415873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,6144,0.012259555359681448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,7168,0.0436426666047838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,6144,0.01904266741540697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,5120,0.01107022249036365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,5120,0.018055111169815063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,6144,0.04038933250639174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,4096,0.01017955525053872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,5120,0.039150221480263606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,4096,0.01707466608948178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,3584,0.009331555830107795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,3584,0.016764443781640794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,3072,0.009192888935407003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,4096,0.03635111120012071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,3584,0.03478399912516276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,3072,0.015136000182893542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,2560,0.007300444775157505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,3072,0.031911111540264554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,2560,0.0148062225845125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,65536,0.058006220393710665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,2048,0.0064177778032090925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,2560,0.03125422199567159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,2048,0.014304889572991265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,1536,0.004974222017659081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,2048,0.02886755598915948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,1536,0.013960000541475085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,1024,0.0041893331540955436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,65536,0.13519910971323648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,1536,0.02861777941385905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,1024,0.01351111133893331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,768,0.0037902221083641052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,1024,0.027618666489919026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,768,0.013360888593726687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,512,0.0034808889031410217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,768,0.026894221703211468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,512,0.013187555803192986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,256,0.0031644445326593188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,512,0.026317333181699116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,256,0.012714666624863943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,128,0.0029271110478374693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,256,0.026087111896938745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,128,0.012732444538010491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,64,0.002789333255754577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,32,0.002817777709828483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,128,0.025602666868103877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,64,0.013173333472675748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,32,0.012268444730175866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,65536,0.075472891330719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,16384,0.01719111038578881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,16384,0.02404444416364034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,12288,0.014288889037238227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,12288,0.021720889541837905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,16384,0.05136977632840475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,10240,0.013048888908492194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,12288,0.049123555421829224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,10240,0.020870221985710993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,8192,0.011409777734014722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,10240,0.046457777420679726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,8192,0.01960533360640208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,7168,0.010762666662534079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,8192,0.043008890416887074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,7168,0.019138667318556044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,6144,0.009783999787436591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,7168,0.04128799835840861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,6144,0.018005333013004728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,5120,0.009400000174840292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,65536,0.052820444107055664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,5120,0.016142222616407607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,6144,0.03918044434653388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,4096,0.008802666432327693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,5120,0.03747377792994181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,4096,0.015461333923869662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,3584,0.00866844422287411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,4096,0.032688889238569475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,3584,0.015463999576038785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,3072,0.008185777399275038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,2560,0.00673333348499404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,3584,0.032376001278559365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,65536,0.11879111660851373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,3072,0.015066666735543145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,3072,0.03202933404180739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,2560,0.014826666977670459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,2048,0.005991111199061076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,2048,0.01440622243616316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,2560,0.0302782224284278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,1536,0.004827555682924059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,2048,0.02844266759024726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,1536,0.013946667313575745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,1024,0.004146666576464971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,1536,0.028155555327733357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,768,0.013404444687896304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,512,0.0034764446318149567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,1024,0.01367111090156767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,768,0.003773333297835456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,1024,0.026752000053723652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,65536,0.050864888562096484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,256,0.003145777723855443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,768,0.02667466633849674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,512,0.013227555486891003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,128,0.012866666747464074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,256,0.012721777790122561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,512,0.02604888876279195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,128,0.0029440000653266907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,256,0.025761778155962627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,64,0.002776888923512565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,32,0.0028648889727062653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,128,0.02555022140343984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,64,0.013157332936922709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,32,0.012281777958075205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,16384,0.014192000031471252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,16384,0.02286311156219906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,12288,0.011653333074516721
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,16384,0.04720888866318596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,12288,0.021009777983029682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,10240,0.011447110937701331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,10240,0.020133333073721994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,8192,0.010267555713653564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,12288,0.04615022076500786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,10240,0.04359199934535556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,8192,0.01851733360025618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,8192,0.04085510969161987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,7168,0.009416888985368941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,7168,0.018226666582955253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,6144,0.008728889127572378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,7168,0.03872266742918227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,6144,0.016626665989557903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,65536,0.048884444766574435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,5120,0.008858666651778752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,6144,0.034987555609809026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,5120,0.016207999653286405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,4096,0.007618666523032718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,65536,0.10992266734441121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,5120,0.03376977642377218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,4096,0.01568000018596649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,3584,0.008085333638721043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,4096,0.03165244393878513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,3584,0.015481778317027621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,3072,0.007191999918884701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,3584,0.030998223357730444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,3072,0.015115555789735583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,2560,0.006257777826653586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,2560,0.014811555544535318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,3072,0.030193779203626845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,2048,0.005501333210203383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,65536,0.0400257772869534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,2048,0.028139554791980322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,2560,0.029548443026012842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,2048,0.014398222168286642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,1536,0.004855999930037392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,1024,0.013561777770519257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,1536,0.0140684445699056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,1024,0.0041573333243529005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,1536,0.027448000179396734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,768,0.0037600000699361167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,1024,0.02660977840423584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,768,0.013356444736321768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,512,0.0034906665484110513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,768,0.026525333523750305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,256,0.012663111090660095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,512,0.013165333204799227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,256,0.0031724443866146933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,512,0.025868443979157343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,128,0.002974222310715251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,128,0.025421332981851365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,256,0.025581333372328017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,128,0.012551999754375882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,64,0.0028080000645584534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,32,0.0028115556471877625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,64,0.013068444199032254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,32,0.012431999875439538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,16384,0.011551111108726926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,16384,0.02270222206910451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,12288,0.01017155581050449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,12288,0.01957244508796268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,16384,0.0454293323887719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,10240,0.009852444132169088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,12288,0.04295288854175144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,10240,0.01853155593077342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,8192,0.008983111215962304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,8192,0.018365333477656048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,7168,0.00849511143234041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,10240,0.03884622123506334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,65536,0.04748622245258755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,8192,0.036488887336519026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,65536,0.09674755732218425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,6144,0.007507555186748505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,7168,0.017086222767829895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,6144,0.016509332590632968
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,7168,0.036080890231662326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,5120,0.007674666742483775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,65536,0.028785778416527644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,4096,0.01568533314598931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,6144,0.0336097776889801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,5120,0.015997333659066092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,4096,0.006771555377377405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,5120,0.032650666104422674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,3584,0.0074622225430276655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,4096,0.031256887647840716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,3584,0.015262222952312894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,3072,0.00720977783203125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,3072,0.01499466598033905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,3584,0.030582222673628066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,2560,0.006153777655627992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,3072,0.029790222644805908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,2560,0.014517333772447375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,2048,0.005526222288608551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,2560,0.029305779271655615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,2048,0.014416888356208801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,1536,0.004775111046102312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,2048,0.028026666906144884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,1536,0.014010666145218743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,1024,0.0041537777417235905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,1536,0.027200889256265428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,768,0.0037875556283526947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,1024,0.013493333425786761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,1024,0.026736888620588515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,768,0.013387555877367655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,512,0.0035004446076022256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,768,0.026421333352724712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,256,0.012671111358536614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,512,0.013145777914259167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,256,0.0031591111587153543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,512,0.025832888152864244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,128,0.002966222249799304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,256,0.025605332520272996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,128,0.012488889197508493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,64,0.002797333316670524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,128,0.025450666745503742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,32,0.002830222249031067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,64,0.013131555583741931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,32,0.012431999875439538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,16384,0.008911111288600499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,65536,0.04632888899909126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,16384,0.021910222040282354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,65536,0.09485422240363227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,12288,0.008191111187140146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,16384,0.04315377937422859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,12288,0.019149333238601685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,10240,0.007384888827800751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,12288,0.04061511158943176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,65536,0.018132444885041978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,8192,0.017338666650984023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,10240,0.018326222896575928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,8192,0.007286222444640265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,10240,0.03834755553139581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,7168,0.006955555743641323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,8192,0.035615109735065036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,7168,0.016864000095261466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,6144,0.006783111227883234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,7168,0.034916443957222834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,6144,0.01647644407219357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,4096,0.006657777975002925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,5120,0.007098666495747036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,6144,0.03336711062325372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,5120,0.015976000163290236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,5120,0.032596445745891996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,3584,0.007142222589916653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,4096,0.015486222174432544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,3584,0.015308444698651632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,4096,0.030863109562132094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,3072,0.006798222247097228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,3584,0.0302471121152242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,2560,0.006168889088763132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,3072,0.014870221416155497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,3072,0.029695110188590154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,2560,0.01458755632241567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,2048,0.005492444253630108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,2560,0.028993778758578833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,2048,0.014177777700954013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,1536,0.004794666750563515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,1536,0.013934221532609729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,2048,0.02770044406255086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,1024,0.004117333226733738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,1536,0.02716000046994951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,768,0.00378577783703804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,1024,0.013567110730542077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,1024,0.02695111102528042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,768,0.013250666360060373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,512,0.0034622223012977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,768,0.026103999879625108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,512,0.013099555340078143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,256,0.0031226666437255014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,512,0.025721778472264607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,128,0.002895999906791581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,256,0.012703110774358114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,256,0.025992888543340895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,64,0.0027262222849660446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,128,0.012495111260149213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,32,0.0027377777215507296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,128,0.025428444147109985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,64,0.013068444199032254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,32,0.012264889147546558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,65536,0.012350222302807702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,16384,0.007102222078376346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,65536,0.045688887437184654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,16384,0.020844444632530212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,65536,0.09400177664226955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,12288,0.006675555474228329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,16384,0.042605333858066134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,12288,0.018980445133315194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,10240,0.006829333388143116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,12288,0.0403733319706387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,10240,0.01796888808409373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,8192,0.006599999964237213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,10240,0.03791911072201199
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,8192,0.017280000779363845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,7168,0.006597333484225803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,8192,0.03594311078389486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,7168,0.016662221815851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,6144,0.006543111056089401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,7168,0.034684442811542086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,6144,0.0162453336848153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,5120,0.00673333348499404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,6144,0.03326933251486884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,5120,0.01568266749382019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,4096,0.00638755535085996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,5120,0.03221599923239814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,4096,0.015341333217091031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,3584,0.006795555353164673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,4096,0.030827555391523574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,3584,0.015088000231319003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,3072,0.00655377780397733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,3584,0.030165334542592365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,3072,0.01458577811717987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,2560,0.005774222314357758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,3072,0.029919998513327703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,2560,0.014429334137174817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,2048,0.005231110999981563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,2560,0.028907554017172918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,1536,0.013879111243618859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,2048,0.014021333720948962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,1536,0.004601777841647466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,2048,0.027620444695154827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,1024,0.003951111187537511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,1024,0.02696177860101064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,1536,0.027278222971492346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,1024,0.013433777623706393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,768,0.0036515556275844574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,768,0.013170666992664337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,512,0.003367111086845398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,768,0.02597155504756504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,512,0.013250666360060373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,256,0.003077333379122946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,512,0.02601066728432973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,256,0.012497777740160624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,128,0.0029084444459941653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,256,0.025555556019147236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,128,0.01255644443962309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,64,0.0027475555737813315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,32,0.002740444408522712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,128,0.025839999318122864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,16384,0.006603555546866522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,64,0.013045333325862885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,32,0.012230222423871359
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,65536,0.009944000177913243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,10240,0.006699555450015598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,65536,0.044834666781955294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,12288,0.006636444479227066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,16384,0.020591111646758187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,12288,0.018760000665982563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,8192,0.00654755574133661
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,10240,0.017828444639841717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,7168,0.0064222220745351576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,8192,0.01696622206105126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,6144,0.006335999816656113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,7168,0.016632889707883198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,5120,0.006685333533419504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,6144,0.016061334146393668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,4096,0.006305777778228124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,5120,0.015491555134455362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,3584,0.006804444309737947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,4096,0.015136889285511441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,3072,0.00646577775478363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,3584,0.014889778362380134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,2560,0.005757333503829108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,3072,0.014615111880832247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,2560,0.014167111780908374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,2048,0.005150222116046482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,1536,0.004540444248252445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,2048,0.013925333817799887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,1024,0.003922666526503033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,1536,0.013814222481515674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,1024,0.013336000343163809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,768,0.003670222229427762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,512,0.003302222324742211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,768,0.01313333296113544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,512,0.012835555606418185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,256,0.0030177777840031516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,128,0.0028604444944196274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,256,0.012544000314341651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,64,0.0027297778675953546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,128,0.012486221889654795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,65536,0.04385777645640903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,32,0.0027217778066794076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,64,0.013036444783210754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,32,0.012139555480745105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,65536,0.00867377801073922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,16384,0.00656088896923595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,12288,0.006551110910044775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,16384,0.020706666840447318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,10240,0.006656888872385025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,12288,0.019122666782803006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,8192,0.006514666808976068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,10240,0.01789066692193349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,7168,0.0064186664919058485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,8192,0.017044444878896076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,6144,0.006308444258239534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,7168,0.016629333297411602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,5120,0.00674933319290479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,6144,0.016201777590645682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,4096,0.006280000011126201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,5120,0.015560888581805758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,3584,0.0068195557428730865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,4096,0.015040000279744467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,3072,0.006416000011894438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,3584,0.014891554911931356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,2560,0.0057884446448749965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,3072,0.014558222558763293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,2048,0.005130666825506422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,2560,0.013996443814701505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,1536,0.004550222307443619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,2048,0.01388533330625958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,1024,0.003946666502290302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,1536,0.013645333548386892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,768,0.003588444242874781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,1024,0.013387555877367655
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,512,0.00332622230052948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,768,0.013037333057986366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,256,0.003014222201373842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,512,0.012901333471139273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,128,0.002784888984428512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,256,0.012612444659074148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,64,0.0026844443960322272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,128,0.012681778106424542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,32,0.0027031112048361036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,64,0.01297511077589459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,32,0.012112889024946423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,10240,0.23410932223002115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,12288,0.2781475649939643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,16384,0.3656328784094916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,12288,0.34653955035739475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,16384,0.41905776659647626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,8192,0.1906426615185208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,10240,0.47334935930040145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,8192,0.24102756712171766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,10240,0.2986622121598986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,12288,0.5685564676920573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,7168,0.16870310571458605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,16384,0.7442000177171496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,8192,0.3779217667049832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,7168,0.2166559961107042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,7168,0.3292897807227241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,6144,0.1473955578274197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,5120,0.12567199601067439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,6144,0.19027466244167754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,4096,0.11836000283559163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,5120,0.23727732234530977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,5120,0.1629706621170044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,4096,0.13497243987189397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,3584,0.09425333473417495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,6144,0.28465867042541504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,3584,0.12186666329701741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,4096,0.1914728879928589
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,3584,0.1671182182100084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,3072,0.08250666989220513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,3072,0.14395288626352945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,2560,0.07206933365927802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,3072,0.10732977920108372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,2560,0.09204977750778198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,2048,0.06081155273649427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,1536,0.05084800057941013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,1024,0.04355022311210632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,2048,0.0757182240486145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,1024,0.05051999953058031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,1536,0.07450489203135173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,1536,0.06466577450434367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,2560,0.12095822228325738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,2048,0.09663377867804633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,1024,0.053598221805360585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,512,0.03224800030390421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,768,0.03678133421474033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,512,0.04251733422279358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,768,0.047713776429494224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,256,0.02864266766442193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,768,0.03934755590226915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,128,0.009214222431182861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,512,0.02725244396262699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,64,0.006487999939256244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,256,0.03580355644226074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,128,0.027225777506828308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,256,0.01533422205183241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,32,0.004573333180612988
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,64,0.023108444280094568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,128,0.033194667763180204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,32,0.02254844374126858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,12288,0.08298222223917644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,16384,0.10454133484098647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,16384,0.13223644097646078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,12288,0.13983110586802164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,12288,0.11029599772559272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,10240,0.07155022356245253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,10240,0.09784177939097087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,16384,0.1893519957860311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,8192,0.0604719983206855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,8192,0.09436800082524617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,10240,0.11637777752346462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,8192,0.07704710960388184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,7168,0.055007110039393105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,7168,0.0738275580936008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,6144,0.049599998527103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,6144,0.06613955232832167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,7168,0.08306755622227986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,5120,0.044093334012561373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,6144,0.07215822405285306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,5120,0.06021866533491346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,5120,0.0602417786916097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,4096,0.03827111257447137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,4096,0.05604622099134657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,4096,0.04900622367858887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,3584,0.03602222270435757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,3584,0.05061155557632446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,3072,0.03351377778583103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,3584,0.05056088831689623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,3072,0.037788444095187716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,3072,0.04776533444722494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,2560,0.03052977720896403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,2560,0.03197777933544583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,65536,0.37223998705546063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,2048,0.02760444415940179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,2048,0.02609066665172577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,2560,0.04483200112978617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,2048,0.041399998797310725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,1536,0.025468443830808003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,1536,0.03730577892727322
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,1024,0.02124622298611535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,1536,0.020647999313142564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,768,0.011678222152921887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,1024,0.03415822320514255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,768,0.019725332657496136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,512,0.008753777378135258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,768,0.03311289019054837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,512,0.01829866733815935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,256,0.00599911105301645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,1024,0.01462933255566491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,512,0.03135022189882066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,256,0.014467555615637036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,128,0.003593777616818746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,256,0.0270142224099901
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,128,0.01499644418557485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,64,0.0032275555034478507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,128,0.026943110757403906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,32,0.003248888792263137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,64,0.014260444376203748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,32,0.013573333621025085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,65536,0.4278435442182753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,16384,0.08573244677649604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,16384,0.11378578344980876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,12288,0.06732266479068331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,16384,0.14496178097195095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,12288,0.09155555566151936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,12288,0.11177244451310898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,10240,0.0888942215177748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,10240,0.05856000052558052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,10240,0.08089244365692139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,8192,0.04996622271007962
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,65536,0.28552267286512584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,8192,0.07066222031911214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,7168,0.045924444993337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,8192,0.07183644506666395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,7168,0.06649511390262179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,7168,0.06342577934265137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,6144,0.04167644513977898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,65536,0.3676266670227051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,6144,0.062021334966023765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,6144,0.05884977844026354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,5120,0.03758666581577725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,5120,0.0558097759882609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,5120,0.04619466596179538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,4096,0.03368533319897122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,4096,0.03748355640305413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,3584,0.031133333841959637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,4096,0.04926310976346334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,3584,0.049350221951802574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,3072,0.03141955534617106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,2560,0.026872888207435608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,3584,0.03305600086847941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,65536,0.740997314453125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,3072,0.0447271102004581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,2560,0.04209688968128628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,2048,0.024338665935728285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,1536,0.02236088944805993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,2048,0.03878488805558946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,3072,0.028728001647525366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,2560,0.024847999215126038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,1024,0.011808888779746162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,1536,0.035798221826553345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,768,0.009766221874290043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,1024,0.03369599911901686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,1536,0.015947555502255756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,1024,0.019899555378490023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,768,0.01910044418440925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,2048,0.02029511166943444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,512,0.007603555917739868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,768,0.03194400005870395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,256,0.004473777694834603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,512,0.017539555827776592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,512,0.0306311117278205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,256,0.014304000470373364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,128,0.003650666524966558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,256,0.02642755541536543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,128,0.014644443988800049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,64,0.0033173333439562055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,32,0.013286222186353473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,128,0.026572444372706946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,64,0.014000889327791003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,32,0.003338666632771492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,16384,0.0735644433233473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,16384,0.09634221924675836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,12288,0.0898044440481398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,12288,0.05936800109015571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,16384,0.1176533301671346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,12288,0.07677421967188518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,10240,0.05230577786763509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,10240,0.06951022148132324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,8192,0.060773332913716636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,10240,0.07481244537565443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,8192,0.0449848903550042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,7168,0.041430221663581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,8192,0.06164444155163235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,65536,0.5541582107543945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,7168,0.05341066585646736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,7168,0.05730488565233019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,6144,0.03801066676775614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,6144,0.054076443115870156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,6144,0.04692622356944614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,65536,0.24758489926656088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,5120,0.03942577706442939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,5120,0.034513778156704374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,4096,0.03106400039460924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,5120,0.050029334094789296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,4096,0.04853511187765333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,3584,0.029363556040657893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,3072,0.027172444595230952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,4096,0.03656889001528422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,3584,0.04418755571047465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,3584,0.028466665082507666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,3072,0.04073333409097459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,2560,0.024959110551410254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,2048,0.023286221755875483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,2560,0.03855289022127787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,3072,0.024892444411913555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,2048,0.03590044379234314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,1536,0.013811555173661975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,1536,0.021372444099850122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,2048,0.017474666237831116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,1024,0.01017066670788659
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,1536,0.03476444549030728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,2560,0.021213332811991375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,1024,0.019354666272799175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,65536,0.312903987036811
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,768,0.008561777571837107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,1024,0.03227555420663621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,768,0.017123555143674214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,512,0.006806222101052602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,768,0.03145422206984626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,512,0.015052444405025907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,256,0.0037271111375755737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,256,0.014137778017255994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,128,0.014599111345079211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,64,0.0030951110853089225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,128,0.0033200000309281876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,512,0.03073955575625102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,256,0.02648799949222141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,128,0.02685333291689555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,32,0.003108444520168834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,64,0.01383733335468504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,32,0.013197333448463015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,16384,0.06086844205856323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,16384,0.08122755421532525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,12288,0.07068444622887506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,65536,0.45840268664889866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,12288,0.05006577902370029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,12288,0.06945955753326416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,16384,0.09220444493823582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,10240,0.048913776874542236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,10240,0.05905244747797648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,8192,0.038422223594453596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,10240,0.06252444452709623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,8192,0.05522400140762329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,8192,0.04785155587726169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,7168,0.042416890462239586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,7168,0.035579555564456515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,6144,0.03275022241804335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,7168,0.05237244566281637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,6144,0.03745333353678385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,6144,0.0503786669837104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,5120,0.029810667037963867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,5120,0.04616444309552511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,5120,0.03142222099834018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,4096,0.026331555512216356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,3584,0.02440266642305586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,4096,0.04232266545295715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,3584,0.043004443248113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,4096,0.02570844524436527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,3072,0.02279377811484867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,3072,0.037984887758890785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,65536,0.19668177763621011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,2560,0.021377777059872944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,3584,0.02290488945113288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,2560,0.036102222071753606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,2048,0.02001066671477424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,2560,0.017440888616773818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,1536,0.011598221957683563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,3072,0.020278222031063504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,1536,0.01850311126973894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,2048,0.03467822074890137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,1024,0.008728000024954477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,1536,0.03388088941574097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,1024,0.01699555582470364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,768,0.007334222396214803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,2048,0.014426666829321118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,1024,0.03131822082731459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,768,0.015544000599119397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,512,0.00609688874748018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,512,0.013597333596812354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,256,0.0035013332962989807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,768,0.029359999630186293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,512,0.027179555760489568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,256,0.012873777912722694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,128,0.0031742221779293488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,256,0.026526222626368206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,128,0.013335111240545908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,64,0.003077333379122946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,32,0.0030862221287356485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,128,0.026632888449562922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,64,0.012679999901188744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,32,0.011979555918110741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,65536,0.2517022291819255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,16384,0.05595733059777153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,16384,0.08005511098437838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,16384,0.08454844686720107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,12288,0.04619733492533366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,12288,0.06562044223149617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,12288,0.06643822458055285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,10240,0.041336887412601046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,10240,0.05552622344758776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,10240,0.060119999779595264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,8192,0.036144889063305326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,8192,0.04591466652022468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,8192,0.056186669402652316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,7168,0.03352800011634827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,7168,0.04180622100830078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,65536,0.37383466296725804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,6144,0.03091199861632453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,7168,0.0506568882200453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,65536,0.17672533459133574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,6144,0.04838755395677355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,6144,0.03600355651643541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,65536,0.20701421631707084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,5120,0.02811466654141744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,4096,0.024516445067193773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,3584,0.02310577697224087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,4096,0.040552000204722084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,5120,0.03512977891498142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,5120,0.04454844527774387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,3584,0.039846221605936684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,3072,0.02161866592036353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,2560,0.020615999897321064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,3072,0.03722666700681051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,4096,0.02365066607793172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,3584,0.0207004447778066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,2560,0.035331557194391884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,2048,0.013231111069520315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,2048,0.019367999500698514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,1536,0.010847110715177325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,1536,0.017863111363516915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,1024,0.008192000289758047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,2048,0.03494488861825731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,2560,0.015819556183285184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,1536,0.0333555539449056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,3072,0.018431110514534842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,768,0.007027555671003129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,768,0.014933332800865173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,1024,0.01716444392999013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,1024,0.03103289008140564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,512,0.005741333381997214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,256,0.003506666670242945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,256,0.012811555630630918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,768,0.02952622373898824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,512,0.013230221966902414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,512,0.027127110295825537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,128,0.0032097777972618737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,256,0.026009778181711834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,128,0.013423999978436364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,64,0.0031582222630580268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,128,0.026563555002212524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,32,0.003015999992688497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,64,0.012626666989591388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,32,0.011940444509188334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,16384,0.05231555634074741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,16384,0.07431111070844862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,12288,0.057887997892167836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,16384,0.07556355661816068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,12288,0.04336622357368469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,12288,0.06330133146709867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,10240,0.038145777251985334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,10240,0.049067556858062744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,10240,0.058083554108937584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,8192,0.03376444511943393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,8192,0.05289333396487766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,65536,0.3214133315616184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,7168,0.031585776143603854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,8192,0.0389013323518965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,7168,0.049294223388036094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,7168,0.03433422247568766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,6144,0.029261334074868098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,5120,0.026464000344276428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,6144,0.0465848876370324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,5120,0.043637331989076406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,5120,0.02529688841766781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,65536,0.15586045053270128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,4096,0.02316444449954563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,6144,0.03444177905718485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,3584,0.02216088937388526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,4096,0.03918044434653388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,3584,0.039451556073294744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,3072,0.021065778202480737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,3072,0.03602844476699829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,2560,0.014666666587193808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,4096,0.02161866592036353
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,3584,0.019366222951147292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,2560,0.019722667005327012
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,3072,0.016913778252071805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,2048,0.012599999705950419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,2560,0.03497155507405599
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,2048,0.01873333255449931
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,1536,0.010112889111042023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,2048,0.03401599989997016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,1536,0.01739911072784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,1024,0.007649777664078607
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,1024,0.014700444208251106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,768,0.006581333362393909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,1536,0.033412443266974554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,1024,0.030375109778510198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,512,0.004693333473470476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,768,0.013818666338920593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,512,0.013220444321632385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,65536,0.20923822455936006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,256,0.003540444291300244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,256,0.013005333642164866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,768,0.029378665818108454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,512,0.02681511143843333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,128,0.003173333282272021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,256,0.026224000586403742
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,64,0.0029297777348094513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,128,0.013271110753218332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,32,0.0029582221888833572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,128,0.026927999324268762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,64,0.01254755589697096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,32,0.011903111305501727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,16384,0.046089778343836464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,16384,0.07238844368192884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,65536,0.28061866760253906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,12288,0.03855289022127787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,12288,0.06166578001446194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,16384,0.06109955575731066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,12288,0.046219554212358266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,10240,0.039015110996034406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,10240,0.034882667991850115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,8192,0.03134844369358487
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,10240,0.05572266711129082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,8192,0.03233777814441257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,8192,0.05112089051140679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,65536,0.13644444942474365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,7168,0.029318223396937054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,7168,0.04860533277193705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,6144,0.02677244444688161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,5120,0.024303999212053087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,6144,0.04474577638838026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,7168,0.028031110763549805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,4096,0.022075555390781824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,5120,0.042133334610197276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,3584,0.02085777786042955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,4096,0.03819466630617777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,4096,0.017630222770902846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,5120,0.021361778179804485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,6144,0.02871111035346985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,3072,0.014031999640994601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,3072,0.019966221517986722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,3584,0.03835466504096985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,2560,0.012313777373896705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,2560,0.01905333333545261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,3584,0.015841777125994366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,3072,0.03586133321126302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,2048,0.010414222048388587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,2560,0.03459466828240289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,2048,0.0181351105372111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,1536,0.008732444710201686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,2048,0.033547557062572904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,1536,0.017229333519935608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,1024,0.00666933341158761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,1536,0.03275377882851495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,1024,0.01405244403415256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,768,0.00609422226746877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,1024,0.0284186667866177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,768,0.01366488883892695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,512,0.0041857775714662345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,768,0.027433777848879497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,512,0.013285333083735572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,256,0.0032844444115956626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,512,0.02674666709370083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,256,0.01277599980433782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,128,0.0031617778456873367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,256,0.02608977754910787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,128,0.01331555595000585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,64,0.0028782222006056043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,65536,0.18845155504014757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,128,0.026547556122144062
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,32,0.0029226665695508323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,64,0.012489777472284106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,32,0.011774222056070963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,16384,0.040579554107454084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,16384,0.06780266761779785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,16384,0.05063199996948242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,12288,0.033936000532574125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,65536,0.23458666271633574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,12288,0.05798844496409098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,12288,0.03977866636382209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,10240,0.031876444816589355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,10240,0.03383466601371765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,10240,0.052413334449132286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,8192,0.02805244591501024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,7168,0.025881777207056682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,8192,0.047163556019465126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,7168,0.04531911015510559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,8192,0.028100444210900202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,6144,0.023938667443063524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,65536,0.11221510834164089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,5120,0.02201777696609497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,7168,0.025415110919210646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,6144,0.04251911242802938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,5120,0.040525333748923406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,4096,0.020657777786254883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,6144,0.02273511058754391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,4096,0.03667466508017646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,3584,0.013228444589508904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,3584,0.019696000549528334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,3072,0.011900444825490316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,5120,0.02012622190846337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,4096,0.016819554898473952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,3584,0.03831200136078729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,3072,0.018960000740157235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,2560,0.010464000205198923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,2560,0.018408000469207764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,2048,0.00907022257645925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,65536,0.186445328924391
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,3072,0.034834666384591
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,2048,0.016721778445773654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,2560,0.03400444322162204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,1536,0.014800000521871777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,1536,0.007680888805124495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,2048,0.033153778976864286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,1024,0.0063991112013657885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,1024,0.013751111096805997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,768,0.004761777818202972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,1536,0.032474666833877563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,768,0.013629333012633853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,1024,0.028408888313505385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,512,0.003644444462325838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,768,0.02695377833313412
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,512,0.013004444539546967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,256,0.003314666656984223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,512,0.02663999961482154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,256,0.012732444538010491
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,128,0.003128000017669466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,256,0.025988444685935974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,128,0.013201778133710226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,64,0.0028728888266616394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,32,0.002934222213096089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,128,0.026557332939571802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,64,0.012304888831244575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,32,0.011734222372372946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,16384,0.03747289048300849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,16384,0.06617244746949938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,12288,0.035820444424947105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,16384,0.04545422063933479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,12288,0.03204355637232462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,12288,0.05681600173314413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,65536,0.1887822283638848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,10240,0.03064089020093282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,8192,0.027295110954178706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,10240,0.02957688768704732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,8192,0.04682577649752299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,10240,0.051924443907207914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,8192,0.02551022171974182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,7168,0.02408444384733836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,7168,0.045075555642445884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,6144,0.022699554761250813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,6144,0.042599111795425415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,5120,0.021031111478805542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,4096,0.013363555901580386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,5120,0.040510220660103693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,7168,0.02365422248840332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,5120,0.018266666266653273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,6144,0.020541333489947848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,4096,0.019867555962668527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,3584,0.012296888563368054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,3584,0.01920000049802992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,4096,0.03678311241997613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,3072,0.010985777609878115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,3584,0.03757866554790073
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,65536,0.1000622246000502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,3072,0.018428444862365723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,2560,0.009638222555319468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,3072,0.03468888998031616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,2560,0.017829333742459614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,2048,0.008466666771305932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,2560,0.033675554725858904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,2048,0.015571556157535978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,1536,0.007119110888904995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,2048,0.032602667808532715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,1536,0.014246222045686511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,1024,0.006036444256703059
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,1536,0.032242665688196816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,1024,0.013749333719412485
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,768,0.003995555556482739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,768,0.013538666897349887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,1024,0.02794755498568217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,512,0.003642666671011183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,65536,0.17184178034464517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,768,0.026945778065257605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,512,0.012905778156386482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,256,0.0033155555526415506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,512,0.026871111657884385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,256,0.012648888760142855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,128,0.0030053332448005676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,128,0.013006221916940479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,256,0.02583733366595374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,64,0.0029084444459941653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,32,0.0029013332807355454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,64,0.012427555190192329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,128,0.02658577760060628
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,32,0.011559999651379056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,65536,0.16443022092183432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,16384,0.034585777256223894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,16384,0.06454488966200086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,16384,0.04047200083732605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,12288,0.029655112160576716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,12288,0.055071999629338585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,10240,0.027051554785834417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,12288,0.03578844335344102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,10240,0.05097777644793192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,8192,0.02425244450569153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,10240,0.02754666739039951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,8192,0.04615644282764859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,7168,0.022853333089086745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,7168,0.04408266809251574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,6144,0.02162133322821723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,7168,0.02075199948416816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,8192,0.022631110416518316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,5120,0.016378665963808697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,6144,0.041851556963390775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,5120,0.020046222541067336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,6144,0.018360000517633226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,4096,0.014895111322402954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,5120,0.039883553981781006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,4096,0.01943644384543101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,4096,0.03611110978656345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,3584,0.013604444762070974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,65536,0.0852666695912679
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,3584,0.01861422260602315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,3072,0.012409778104888068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,3584,0.037420445018344455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,3072,0.01814399990770552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,2560,0.009112888740168678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,2560,0.017128000656763714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,3072,0.03446133269204034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,2048,0.008032889001899296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,2560,0.033560888634787664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,2048,0.015016888578732809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,1536,0.006871999965773688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,2048,0.03277777632077535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,1536,0.014170666535695394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,1024,0.005527110977305307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,1536,0.03007555670208401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,1024,0.01370666672786077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,65536,0.15861599975162083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,768,0.004401777767472797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,1024,0.027344889110989038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,768,0.013371555341614617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,512,0.003645333151022593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,768,0.027042667071024578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,512,0.012949333422713809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,256,0.0032888888898822996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,256,0.025964443882306416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,512,0.026473777161704168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,256,0.01257333325015174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,128,0.003042666655447748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,128,0.013165333204799227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,64,0.0028808888875775864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,32,0.002895111011134254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,64,0.012433778080675336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,128,0.026349332597520616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,32,0.011597333682907952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,16384,0.03108000093036228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,16384,0.06449066930347018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,12288,0.026659554905361597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,12288,0.02645688917901781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,12288,0.05500177873505486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,10240,0.024168888727823894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,65536,0.1439733372794257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,16384,0.03366400135887994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,10240,0.049545778168572314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,8192,0.023583110835817125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,8192,0.04469511244032118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,10240,0.02346933384736379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,7168,0.021331555313534204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,7168,0.04509422183036804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,6144,0.016283555163277518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,7168,0.018206222189797294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,6144,0.020370667179425556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,8192,0.019903999235894945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,65536,0.07598755757013957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,5120,0.01457244488928053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,6144,0.04084711273511251
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,5120,0.019366222951147292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,4096,0.012892444100644855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,4096,0.018518222702874076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,3584,0.011951111257076263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,5120,0.0402693317996131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,4096,0.03561333484119839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,3584,0.018147556318177115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,3072,0.011197333534558615
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,2560,0.008416000339719985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,3584,0.03698755635155572
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,3072,0.017080888152122498
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,2560,0.016407999727461074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,3072,0.03421955638461643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,2560,0.03330311179161072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,2048,0.007142222589916653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,2048,0.014897776974572076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,1536,0.0064088888466358185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,2048,0.030932443009482488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,1536,0.014236445228258768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,1024,0.0047111109726958806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,1024,0.013635555903116861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,1536,0.029988444513744775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,1024,0.02732444471783108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,768,0.003951111187537511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,768,0.02685333291689555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,65536,0.14527377817365858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,768,0.013307555682129331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,512,0.003629333443111844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,512,0.012991111311647626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,256,0.0032346666687064697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,512,0.026408000124825373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,256,0.012827555338541666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,128,0.0029804443733559716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,256,0.025806221697065566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,128,0.013069333301650153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,64,0.0028497777465316984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,128,0.026331555512216356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,32,0.0028577778074476454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,64,0.012506666282812754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,32,0.011626666618718041
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,16384,0.027830223242441814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,16384,0.06265511115392049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,12288,0.023912888434198167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,16384,0.027515555421511333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,10240,0.023060444328520033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,65536,0.12020356125301784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,12288,0.05240977803866068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,12288,0.022600000103314716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,10240,0.04793688986036512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,8192,0.021311110920376245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,8192,0.04413333204057482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,7168,0.015269334117571512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,10240,0.019938665959570143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,7168,0.020451555649439495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,65536,0.06862133079104953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,6144,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,8192,0.01665244499842326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,6144,0.019324445062213473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,5120,0.012771555946932899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,7168,0.04314666655328539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,6144,0.04066310988532172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,5120,0.018830221560266282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,4096,0.011235555840863122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,3584,0.01034311122364468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,4096,0.017867555220921833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,5120,0.038930667771233454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,65536,0.15816356076134577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,3584,0.016619554824299283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,4096,0.03555466731389364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,3072,0.009974222216341231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,3584,0.037407110134760536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,3072,0.015928000211715698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,2560,0.007692444655630324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,3072,0.03388888968361749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,2560,0.015336889359686108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,2048,0.006904888898134232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,2560,0.031096001466115315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,2048,0.01480977733929952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,1536,0.00562755556570159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,2048,0.030180444320042927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,1536,0.014231999715169271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,1024,0.004159111115667555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,1536,0.029642668035295274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,1024,0.01349422252840466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,768,0.0038773334688610504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,1024,0.027753776974148218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,768,0.01332622187005149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,512,0.0035084444615576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,768,0.027062222361564636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,512,0.012860444684823355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,256,0.0031795555518733132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,512,0.026399999856948853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,256,0.012788444757461548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,128,0.0029475556479560006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,256,0.025998221503363714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,128,0.013011555704805585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,64,0.002871111035346985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,128,0.026391999589072332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,32,0.0028213332924577924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,64,0.012455110748608908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,32,0.01166133334239324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,16384,0.025395555628670588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,65536,0.09716800186369155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,16384,0.05944355328877767
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,12288,0.0236453331179089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,12288,0.05154222249984741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,10240,0.016574222180578444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,10240,0.022034666604465906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,16384,0.02275377843115065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,12288,0.01850400037235684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,10240,0.047453333934148155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,8192,0.01425155500570933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,8192,0.020473778247833252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,7168,0.013110222087966071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,8192,0.04358222087224325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,7168,0.019783111082182992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,6144,0.012275555895434486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,7168,0.04221066501405504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,6144,0.01866044435236189
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,65536,0.062218666076660156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,5120,0.011188444164064197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,6144,0.040046221680111356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,5120,0.01831733352608151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,4096,0.010050666828950247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,5120,0.038281778494517006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,4096,0.016671111186345417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,3584,0.009602666729026372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,3584,0.015646222564909194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,4096,0.03485600153605143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,3072,0.009043555292818282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,3584,0.0363031095928616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,65536,0.14709333578745523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,2560,0.014767999450365702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,3072,0.015239111251301236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,2560,0.007339555356237624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,3072,0.030905776553683813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,2048,0.00629333323902554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,2560,0.030181331766976252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,1536,0.004871110949251386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,2048,0.014356444279352823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,2048,0.029304001066419814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,1536,0.014005333185195923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,1024,0.0041706665522522396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,1536,0.02899555696381463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,768,0.013117333253224691
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,1024,0.013527111046844058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,768,0.0037964445849259696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,1024,0.02701955537001292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,512,0.0035111109415690103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,768,0.026756443911128577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,256,0.0032151111712058387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,512,0.012900444368521372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,512,0.026310222016440496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,128,0.002983999955985281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,256,0.012490666574902005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,256,0.025815111067559984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,128,0.013045333325862885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,64,0.0027911110470692315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,32,0.0028613333900769553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,128,0.02637333340115017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,64,0.012322666744391123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,32,0.01145866678820716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,65536,0.07526577843560113
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,16384,0.023785778217845496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,12288,0.014204444156752693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,16384,0.05434222353829277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,12288,0.021739555729760066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,10240,0.012993777791659037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,12288,0.04964088731341892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,10240,0.020460445019933913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,8192,0.01131199962562985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,10240,0.04558311237229241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,16384,0.01723911033736335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,8192,0.019193778435389202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,7168,0.010739555789364709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,7168,0.01899111105336083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,8192,0.041688889265060425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,6144,0.00978133330742518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,7168,0.042303111818101674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,6144,0.017862222260899015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,5120,0.009335999687512716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,65536,0.05154222249984741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,6144,0.03817688756518894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,5120,0.016023111012246873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,4096,0.008800889054934183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,4096,0.015368888775507608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,5120,0.035536001125971474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,3584,0.008649778034951951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,65536,0.12510844071706137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,3584,0.015071999695565967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,4096,0.03181511163711548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,3072,0.008099555969238281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,3584,0.032808889945348106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,3072,0.015021334091822306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,2560,0.006333333336644703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,3072,0.03141511148876614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,2560,0.014733332726690503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,2048,0.005894222193294101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,2560,0.02958577871322632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,2048,0.014426666829321118
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,1536,0.005068444543414646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,1536,0.013863999810483722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,2048,0.028445333242416382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,1024,0.004192000048028098
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,1536,0.028219554159376357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,1024,0.013678222066826291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,65536,0.05046933227115207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,768,0.003825777934657203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,1024,0.026918222506841023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,768,0.013148444394270578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,512,0.0034657776769664553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,768,0.0265608893500434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,512,0.013067555924256643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,256,0.0031475555151700974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,512,0.026124444272783067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,64,0.003117333269781537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,256,0.012631110846996307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,128,0.0029173334025674393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,256,0.025597333908081055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,128,0.013108444710572561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,128,0.026551110876931086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,64,0.012315555579132505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,32,0.0031519999934567344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,32,0.011606222225560082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,16384,0.014468444718254937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,16384,0.022634666826989915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,12288,0.012404444317022959
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,12288,0.020976889464590285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,16384,0.051177776522106595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,10240,0.01148799972401725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,12288,0.04681244492530823
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,10240,0.020005333754751418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,8192,0.010129777921570672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,10240,0.04268444577852885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,8192,0.018198221921920776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,65536,0.046911110480626426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,7168,0.009687999884287516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,8192,0.03976266582806905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,7168,0.01756355497572157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,6144,0.008897778060701158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,7168,0.038078222009870745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,6144,0.016851555969980028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,5120,0.008767111433876885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,65536,0.11532621913486057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,6144,0.03604000144534641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,5120,0.01595288846227858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,65536,0.039651556147469416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,4096,0.007891555627187094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,5120,0.034770667552948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,4096,0.015377778146002026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,3584,0.007663999994595845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,3584,0.015180443723996481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,4096,0.030976000759336684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,3072,0.00702311098575592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,2560,0.014612444572978549
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,3584,0.03204710947142707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,3072,0.01497333414024777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,2560,0.029122667180167303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,2560,0.006172444257471297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,3072,0.02962755494647556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,2048,0.005479111025730769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,1536,0.004824000100294749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,1536,0.028166221247778997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,2048,0.014222222897741528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,2048,0.028351998991436426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,1536,0.014006222287813822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,1024,0.0041128889554076726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,1024,0.01348444405529234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,768,0.0037760001917680106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,1024,0.026541334059503343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,768,0.013304000099500021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,512,0.003488000068399641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,768,0.026299556096394856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,512,0.013092444174819522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,256,0.003391111062632667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,512,0.025967111190160114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,256,0.01260355528857973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,128,0.003100444459252887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,256,0.025674665967623394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,128,0.013062222136391534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,64,0.0029528888149393927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,128,0.02621777852376302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,32,0.0028782222006056043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,64,0.012344000240166983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,32,0.011609777808189392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,16384,0.011700444751315646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,16384,0.0224906669722663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,12288,0.010120000276300643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,16384,0.046125332514444985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,12288,0.02075377768940396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,10240,0.00960977789428499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,12288,0.04212977819972568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,10240,0.019010666343900893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,8192,0.008847110801272923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,10240,0.03917955689960056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,8192,0.017218665944205392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,65536,0.04609866605864631
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,65536,0.10304888751771714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,7168,0.016833777228991192
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,7168,0.008625778059164682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,8192,0.03529511226548089
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,6144,0.007694222033023834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,7168,0.03432800041304694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,6144,0.016368889146380954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,5120,0.007719111111429002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,65536,0.0282008879714542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,6144,0.03247733248604669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,5120,0.01572533282968733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,4096,0.006878222028414409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,5120,0.031698667340808444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,4096,0.015372445185979208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,3584,0.0073831114504072405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,4096,0.029876444074842665
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,3072,0.01479111115137736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,3584,0.014916444818178812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,3072,0.00700800038046307
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,3584,0.03049955434269375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,2560,0.006227555374304454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,3072,0.028775999943415325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,2048,0.014208889669842191
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,2560,0.014556444353527494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,2048,0.02752799954679277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,2048,0.005575110928879843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,2560,0.02815022071202596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,1536,0.004876444323195351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,1536,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,1024,0.004277333203289243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,1536,0.02754488918516371
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,1024,0.01351288871632682
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,512,0.003624000069167879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,768,0.003943110919660992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,1024,0.026305778159035578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,768,0.013275555438465543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,768,0.026343999637497797
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,512,0.012810666528013019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,256,0.003257777748836411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,512,0.025919111238585577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,256,0.012479110724396177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,128,0.003015999992688497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,128,0.026201777988009985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,256,0.0255039996571011
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,128,0.013042666845851474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,64,0.0028417776856157514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,32,0.002806222273243798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,64,0.012327999704413943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,32,0.011569778124491373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,16384,0.008671999805503422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,65536,0.04448622133996752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,16384,0.020991111795107525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,12288,0.008242666721343994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,65536,0.09370399845971002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,16384,0.043509334325790405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,10240,0.007697777615653143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,12288,0.019144889381196763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,8192,0.007163555257850223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,65536,0.017786666750907898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,10240,0.018111111389266122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,12288,0.03934311204486423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,8192,0.017305778132544625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,10240,0.03695111142264472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,7168,0.007063111497296228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,8192,0.03465244505140517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,7168,0.016784888174798753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,6144,0.006664000037643645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,5120,0.01577422188387977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,7168,0.03336177931891547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,6144,0.016184000505341422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,5120,0.007033777733643849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,4096,0.015409777561823526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,3584,0.006974221931563483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,6144,0.032111111614439226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,4096,0.006646222124497096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,5120,0.0314995547135671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,4096,0.029630220598644678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,3072,0.006789333290523953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,3584,0.015034667319721647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,3072,0.014783999986118741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,3584,0.02934044599533081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,2560,0.028083556228213843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,2560,0.006218666831652324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,3072,0.02832533253563775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,2560,0.0143857780430052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,2048,0.005481777919663324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,2048,0.014105778601434497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,1536,0.004823110997676849
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,2048,0.027442667219373915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,1536,0.013742222554153867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,1024,0.004120888809363048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,1536,0.02717155549261305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,1024,0.01329955541425281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,768,0.0038035557501845886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,1024,0.026745777991082933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,768,0.013295110729005603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,512,0.003478222423129611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,768,0.026205332742797002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,512,0.012871999707486896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,256,0.003128000017669466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,512,0.025954667064878676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,256,0.012449777788586087
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,128,0.0029102222373088202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,256,0.02569244470861223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,128,0.01293688929743237
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,64,0.0027875554644399216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,128,0.02608266638384925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,32,0.0027866665687825945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,64,0.01221866657336553
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,32,0.011543999943468304
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,65536,0.012640000217490725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,16384,0.0070062221752272705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,65536,0.04398222102059258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,16384,0.02068444424205356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,12288,0.006622222148709827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,65536,0.09308355384402806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,16384,0.043476445807351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,12288,0.018985778093338013
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,10240,0.006765333314736684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,12288,0.03883466786808438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,10240,0.01794666714138455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,8192,0.006631999793979857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,10240,0.0369777778784434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,8192,0.01696266730626424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,7168,0.006536888993448681
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,8192,0.034705777963002525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,6144,0.006393777827421824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,7168,0.016602666841612924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,6144,0.015951999359660678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,5120,0.015619556109110514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,7168,0.03332711259524027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,5120,0.006774222271309958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,6144,0.03243111239539252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,4096,0.006351999938488007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,4096,0.029721776644388836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,5120,0.031304889255099826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,4096,0.015076445208655464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,3584,0.006787555499209299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,3072,0.014484445254007975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,3584,0.014826666977670459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,3072,0.006480000085300869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,3584,0.029245333539115057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,2560,0.00582755563987626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,3072,0.02828177809715271
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,2560,0.014077333940400017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,2048,0.005190222213665645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,2560,0.028113779094484117
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,2048,0.013982221484184265
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,1536,0.004642666627963384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,2048,0.027385777897304956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,1024,0.004016889052258598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,1536,0.013464888764752282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,1536,0.026955554882685345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,1024,0.013231111069520315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,768,0.0038168889780839286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,1024,0.02645955483118693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,768,0.013004444539546967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,256,0.012489777472284106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,512,0.0034711110509104202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,768,0.026013332936498854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,512,0.012664000193277994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,256,0.003101333354910215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,512,0.025765332910749648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,128,0.0029297777348094513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,256,0.025594666600227356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,128,0.013000888956917657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,64,0.0028026666906144884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,128,0.026463111241658527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,32,0.0027475555737813315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,64,0.012282667060693106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,32,0.011437333292431302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,65536,0.010031999813185798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,16384,0.006515555497672822
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,65536,0.04323200053638882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,12288,0.006495111104514863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,7168,0.006480000085300869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,16384,0.020562666985723708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,12288,0.018522666560279
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,10240,0.006634666687912411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,10240,0.017620444297790527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,8192,0.006528000036875407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,8192,0.01676266723208957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,4096,0.006270222365856171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,6144,0.0063688887490166565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,3584,0.006752000086837345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,7168,0.016379555066426594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,6144,0.015855999456511605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,2560,0.005763555566469829
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,3072,0.014360888136757744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,5120,0.006650666809744305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,2048,0.005176888985766305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,5120,0.015077332655588785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,4096,0.014920888675583733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,1536,0.01366133325629764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,3072,0.006493333313200209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,3584,0.01444355481200748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,2560,0.014152889450391134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,1536,0.004510222209824456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,2048,0.013901333014170328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,1024,0.003910222401221593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,768,0.003653333418899112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,1024,0.013297778036859302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,768,0.013001778059535556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,512,0.0033253334048721525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,256,0.003119999956753519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,512,0.012605333493815528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,32,0.0027742222365405825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,128,0.0029306666304667792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,256,0.01240799989965227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,64,0.002806222273243798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,128,0.012996444271670448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,64,0.012144000165992312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,32,0.011379555695586734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,65536,0.008700444466537898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,65536,0.04234755701488919
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,16384,0.006734222173690796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,12288,0.006551110910044775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,16384,0.020541333489947848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,12288,0.01886133352915446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,10240,0.00658044425977601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,8192,0.00656444455186526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,10240,0.0178195552693473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,6144,0.0159599996275372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,7168,0.006461333483457565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,8192,0.016724444097942777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,6144,0.006341333190600078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,7168,0.016473778420024447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,5120,0.00674044465025266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,4096,0.006241777704821692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,5120,0.01551644504070282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,3584,0.006716444260544247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,4096,0.014922666880819531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,3072,0.006414222220579783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,3584,0.014695111248228284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,2560,0.005746666755941179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,3072,0.014521777629852295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,2048,0.005136888888147143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,2560,0.014178666803571912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,1536,0.004534222185611725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,2048,0.013791999883121915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,1024,0.003921777837806278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,1536,0.013481777575280933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,768,0.003604444364706675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,1024,0.01346933344999949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,512,0.003367111086845398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,768,0.012984889249006907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,256,0.002970666728085942
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,512,0.012648888760142855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,256,0.012448888685968188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,128,0.0028311111446883944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,64,0.0027991111079851785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,128,0.013003555436929068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,64,0.012177777787049612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,32,0.002768888862596618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,32,0.01149511088927587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,10240,0.23250932163662383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,12288,0.27728088696797687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,16384,0.3641200065612793
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,12288,0.3511848979526096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,16384,0.41877155833774143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,10240,0.47251113255818683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,8192,0.18986666202545166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,10240,0.2967386775546604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,12288,0.5685155656602647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,8192,0.24472178353203666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,16384,0.7450355423821343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,7168,0.195560892422994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,8192,0.3782293266720242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,7168,0.2170426713095771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,7168,0.32867733637491864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,6144,0.14773778120676676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,4096,0.11810844474368626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,6144,0.19113777743445504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,5120,0.12467643949720596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,5120,0.1641119983461168
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,5120,0.2796666622161865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,4096,0.19061866071489122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,6144,0.28376976648966473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,4096,0.13788178231981066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,3584,0.09638933340708415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,3072,0.08191200097401936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,3584,0.12330755922529434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,3584,0.16689956188201904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,3072,0.11027199692196316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,2560,0.07195378012127347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,2560,0.08933955430984497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,2048,0.06032533115810818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,2048,0.08522578080495198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,1536,0.05110399921735128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,2560,0.12037866645389134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,3072,0.16937243938446045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,2048,0.09615377585093181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,1024,0.04012889001104567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,1536,0.0665644473499722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,1536,0.07403910822338529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,768,0.037455999188952975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,768,0.049842665592829384
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,1024,0.05527822176615397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,512,0.03255911005867852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,256,0.029179556502236262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,1024,0.050428443484836154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,512,0.027143999934196472
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,128,0.009263111485375298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,256,0.03823288944032457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,64,0.006504888749784893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,128,0.026313778426912095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,256,0.015211555692884656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,128,0.03163644340303209
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,768,0.045351111226611666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,64,0.022359111242824133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,32,0.004598222259018156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,512,0.04496444596184624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,32,0.02255200015174018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,16384,0.10489866468641494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,12288,0.08277422189712524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,16384,0.13614843951331243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,12288,0.14009333981408015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,10240,0.07096355491214328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,12288,0.11107555362913345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,16384,0.18437422646416557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,10240,0.11613689528571235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,10240,0.09771022531721328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,8192,0.06083466609319051
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,8192,0.07738933298322889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,7168,0.054788443777296275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,8192,0.09425689114464654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,7168,0.08261066675186157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,7168,0.07294844256507026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,65536,0.37131555875142414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,6144,0.04941155513127645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,6144,0.06659377945794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,5120,0.044534222947226636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,6144,0.07214666737450494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,5120,0.06015555726157295
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,5120,0.0655919975704617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,4096,0.038247111770841814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,3584,0.03896711270014445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,4096,0.048826667997572154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,4096,0.055171555942959256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,3584,0.05202577842606438
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,3072,0.03312711252106561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,3584,0.04315733247333103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,2560,0.030356443590588037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,3072,0.04918577935960558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,3072,0.037575999895731606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,2048,0.02703911066055298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,2560,0.04644088943799337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,2560,0.0318853325313992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,1536,0.02366488840844896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,1536,0.038968000147077776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,1024,0.02090577781200409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,2048,0.04273777869012621
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,2048,0.025961778230137292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,768,0.011693333586057028
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,768,0.020098666350046795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,1024,0.03579555617438422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,1536,0.02050933407412635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,512,0.008766222331258986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,768,0.03515200151337518
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,256,0.005969777703285217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,1024,0.014527110589875115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,512,0.01830044388771057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,128,0.003527111063400904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,256,0.015031110909250049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,512,0.03292799989382426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,256,0.03010933266745673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,128,0.014270222849316068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,64,0.00318755561278926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,32,0.0031840000301599503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,128,0.025386666258176167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,64,0.013493333425786761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,32,0.013335111240545908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,65536,0.42742133140563965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,16384,0.08343555529912312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,16384,0.11702488528357612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,12288,0.06668800115585327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,16384,0.14444888962639704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,12288,0.09525955385631985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,12288,0.11098666985829671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,10240,0.08745421965916951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,10240,0.05830844243367513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,65536,0.7401030858357748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,8192,0.04952088991800944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,10240,0.08545600043402778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,7168,0.04547199938032362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,8192,0.07025955782996283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,8192,0.08781955639521281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,7168,0.06734844711091784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,7168,0.06437866555319892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,6144,0.04158933295143975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,5120,0.04088977641529507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,65536,0.2858364317152235
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,6144,0.05547555618815952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,6144,0.06200799677107069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,65536,0.3677253458234999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,5120,0.05636799997753567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,5120,0.04559200008710226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,4096,0.03310666812790765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,4096,0.05110933383305868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,4096,0.03756355577045017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,3584,0.033843557039896645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,3584,0.03112444281578064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,3072,0.028951111767027114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,3584,0.04825777808825175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,2560,0.02605955633852217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,3072,0.04785244332419502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,2560,0.04318311148219638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,3072,0.030999110804663763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,2048,0.023761777414215937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,2048,0.0395706660217709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,1536,0.021743999587164983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,1536,0.03718044360478719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,1024,0.011735999749766456
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,1024,0.019725332657496136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,1536,0.01595288846227858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,768,0.009925333162148794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,2048,0.020724443925751578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,2560,0.02498400045765771
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,1024,0.034549332327312894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,768,0.019190222024917603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,512,0.007657777931955125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,512,0.016203555795881484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,256,0.004934222333961063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,768,0.03410044312477112
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,512,0.03199022346072727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,256,0.014783110883500842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,128,0.003560888684458203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,256,0.029022223419613306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,128,0.013845333622561561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,64,0.003334222154484855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,32,0.003301333429084884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,64,0.013241777817408243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,128,0.025218667255507574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,32,0.013199999928474426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,16384,0.09582755300733779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,16384,0.0728071133295695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,12288,0.09015822410583496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,12288,0.05879288911819458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,16384,0.11822844876183404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,12288,0.07851111226611666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,10240,0.05147644546296862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,8192,0.04439822170469496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,10240,0.07058666812049018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,10240,0.07526755332946777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,8192,0.06068355507320828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,8192,0.06129333045747545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,7168,0.041152000427246094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,7168,0.057429333527882896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,7168,0.05354933275116814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,65536,0.5588373078240289
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,6144,0.04685689012209574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,6144,0.03770133190684848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,65536,0.23794311947292754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,5120,0.03754222061898973
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,6144,0.05396711164050632
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,65536,0.31381066640218097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,5120,0.05053777827156914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,5120,0.03965866565704346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,4096,0.0329786671532525
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,4096,0.031792001591788396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,3584,0.028891556792789038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,3072,0.027012444204754297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,4096,0.04719644453790453
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,3584,0.044545776314205594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,2560,0.024668445189793903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,3072,0.0422071119149526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,2560,0.040151112609439425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,3584,0.0284186667866177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,2048,0.02271644439962175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,2048,0.037485334608289934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,1536,0.014183110660976835
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,1536,0.02106399999724494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,1536,0.035953776703940496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,1024,0.010238221950001186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,1024,0.01942488882276747
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,2560,0.0210817770825492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,2048,0.017443555924627516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,3072,0.024857777688238356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,768,0.008402666283978356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,1024,0.03404622276624044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,768,0.018560889694425795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,512,0.00674222202764617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,768,0.03435111045837402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,512,0.029143999020258587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,512,0.014897776974572076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,256,0.003938666648334927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,256,0.014635556273990206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,128,0.0033208889265855155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,256,0.028664890262815688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,128,0.013990221752060784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,64,0.003061333257291052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,32,0.0030666666312350165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,128,0.025387555360794067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,64,0.013183111117945777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,32,0.01298577752378252
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,16384,0.060263997978634305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,16384,0.08641599946551853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,12288,0.049598223633236356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,65536,0.4658666716681586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,16384,0.09224622117148505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,12288,0.06968888971540663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,10240,0.04852622085147434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,12288,0.07030666536755033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,10240,0.06487733125686646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,10240,0.05869955486721463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,8192,0.03779911001523336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,8192,0.05793688694636027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,8192,0.047891553905275136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,7168,0.03512710995144314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,7168,0.05259022116661072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,6144,0.032483554548687406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,7168,0.042295998997158475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,6144,0.05007288853327433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,5120,0.03145600027508206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,65536,0.2535333368513319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,6144,0.036883556180530123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,65536,0.19498222404056123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,5120,0.02976622184117635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,3584,0.02422133253680335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,5120,0.046763555871115796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,4096,0.02574399941497379
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,4096,0.02555022140343984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,3584,0.04101066788037618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,3072,0.02251911163330078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,4096,0.04333422250217862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,2560,0.021320889393488567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,3072,0.03897066579924689
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,2560,0.03733866744571262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,2048,0.019848000672128465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,3584,0.022839999861187402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,1536,0.011663999822404651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,2048,0.014374223020341663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,1536,0.018439999885029264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,2560,0.017313778400421143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,2048,0.03709777858522203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,1024,0.008801778157552084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,3072,0.023253333237436082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,1536,0.03456177645259433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,1024,0.016866667403115165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,768,0.007419555551475949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,1024,0.03313510947757297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,768,0.0332951115237342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,768,0.01609599921438429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,512,0.00602311102880372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,512,0.013239111337396832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,256,0.003587555554178026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,512,0.028958221276601154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,256,0.013336000343163809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,128,0.0032337777730491427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,256,0.028393778536054824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,128,0.012667555775907306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,64,0.002979555477698644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,128,0.02516266703605652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,32,0.003055111194650332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,64,0.011901333100265928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,32,0.011786667009194693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,16384,0.056264003117879234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,16384,0.08233066399892171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,12288,0.0651822222603692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,12288,0.04597244328922696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,12288,0.06696799728605482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,16384,0.08424711227416992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,10240,0.0551457769340939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,10240,0.06267822451061673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,10240,0.04086755712827047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,8192,0.035368889570236206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,8192,0.045435554451412626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,8192,0.0572213331858317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,7168,0.03288622366057502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,65536,0.3733182218339708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,7168,0.040447112586763166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,7168,0.05429955654674106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,6144,0.030348443322711523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,6144,0.03590755661328634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,6144,0.048009776406817965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,5120,0.027854220734702215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,5120,0.04499377806981405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,65536,0.1750506692462497
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,4096,0.024712888730896845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,4096,0.041555556986067034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,65536,0.2302302254570855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,5120,0.03075288732846578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,3584,0.023357333408461675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,4096,0.023266666465335425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,3072,0.02148355543613434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,3584,0.04169422056939867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,3072,0.039149334033330284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,2560,0.020451555649439495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,2048,0.01331822243001726
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,3584,0.02087822225358751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,2560,0.036715557177861534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,2048,0.019154666198624503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,1536,0.01089511149459415
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,2048,0.03526044554180569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,1024,0.008197333249780867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,3072,0.018311111463440787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,1536,0.017779555585649278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,2560,0.015966221690177917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,1024,0.015973332855436537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,1536,0.034099555677837796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,768,0.00703022215101454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,1024,0.03250133328967624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,768,0.014170666535695394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,512,0.005585777676767773
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,768,0.03195200032658047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,512,0.013101333545313941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,256,0.0034862222770849862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,512,0.03029600116941664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,256,0.01332355539004008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,128,0.0032177778581778207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,256,0.028353777196672227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,128,0.012368888490729861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,64,0.0029680000411139596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,32,0.0029884444342719186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,128,0.02495199938615163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,64,0.01180622229973475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,32,0.011728000309732227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,16384,0.05208711160553826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,16384,0.07714044385486178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,16384,0.07554666863547431
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,12288,0.04271999994913737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,12288,0.057224002149369985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,12288,0.06912799676259358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,10240,0.04864355590608385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,10240,0.0383200016286638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,10240,0.06017422013812595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,8192,0.033619556162092425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,65536,0.3222951094309489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,8192,0.03967555695109897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,8192,0.052314665582444936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,7168,0.031405333015653826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,7168,0.035199112362331815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,6144,0.029091556866963703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,65536,0.16197155581580266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,7168,0.05395022365781996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,6144,0.046824001603656344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,5120,0.02626577847533756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,5120,0.04389866524272495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,5120,0.025939555631743536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,4096,0.022978666755888198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,6144,0.030518223841985066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,4096,0.04012889001104567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,3584,0.021892444954978094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,3584,0.03864977757136027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,3072,0.021112889051437378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,3072,0.03772266705830892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,2560,0.014712888333532544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,2560,0.019799111617936026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,4096,0.02145333256986406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,2048,0.012473777764373355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,2560,0.036321779092152916
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,2048,0.01864088906182183
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,3584,0.01906755566596985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,3072,0.017049777838918898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,1536,0.010057777994208867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,2048,0.03494133220778571
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,1024,0.007664889097213745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,1536,0.017380444539917838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,768,0.006517333288987477
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,1024,0.01391999920209249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,1536,0.034068445364634194
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,1024,0.03112799922625224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,768,0.014037332601017423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,512,0.0047564444442590075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,65536,0.21829956107669404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,768,0.03219466739230686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,512,0.013044444223244986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,256,0.003452444449067116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,256,0.028423110644022625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,128,0.012434666355450949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,512,0.028346667687098186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,256,0.01332622187005149
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,128,0.003173333282272021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,64,0.002959111084540685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,128,0.02516266703605652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,32,0.0029724445194005966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,64,0.01180622229973475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,32,0.011662222445011139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,16384,0.045384890503353544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,16384,0.0730453332265218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,12288,0.03804888990190294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,12288,0.046486222081714206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,65536,0.28039201100667316
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,16384,0.0589324434598287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,12288,0.06598933537801106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,10240,0.0344124436378479
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,10240,0.058176000912984215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,10240,0.03926666577657064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,8192,0.031015998787350122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,8192,0.03192266821861267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,8192,0.050536000066333346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,7168,0.02861777941385905
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,7168,0.027956446011861164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,6144,0.026144888665941026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,6144,0.045800887876086764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,5120,0.024088889360427856
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,7168,0.04923199945025974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,5120,0.043656888935301036
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,5120,0.024254222710927326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,4096,0.021904889080259535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,6144,0.025019556283950806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,65536,0.18510044945610893
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,4096,0.03937155670589871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,65536,0.13230221801333958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,4096,0.017587555779351127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,3584,0.02074399921629164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,3072,0.013975110318925647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,3584,0.03808888792991638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,3072,0.019753777318530612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,2560,0.012236444486512078
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,3072,0.03714133302370707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,2560,0.018808888064490426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,2048,0.010458666417333813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,3584,0.015770667129092746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,2560,0.0358568893538581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,2048,0.017788444956143696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,1536,0.008725333544943068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,2048,0.03442222211096022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,1536,0.017113778326246474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,1024,0.0147386673423979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,1024,0.006923555499977536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,1536,0.03365244468053182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,1024,0.03151822090148926
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,768,0.005762666463851929
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,768,0.0140364451540841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,768,0.029785778787400987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,512,0.004102222207519743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,512,0.013008000122176277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,256,0.003432000055909157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,512,0.028135110934575398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,256,0.013405332962671915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,128,0.0030906666070222855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,256,0.028304000695546467
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,128,0.012291555603345236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,64,0.0028702221396896574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,128,0.025024000141355727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,64,0.011884444289737277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,32,0.002926222152180142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,32,0.011612444288200803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,16384,0.04011911153793335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,16384,0.06939733028411865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,12288,0.03385600116517808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,12288,0.03956355651219686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,16384,0.05056533217430115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,12288,0.05962933434380425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,10240,0.03048355711830987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,10240,0.03386133246951633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,10240,0.05537422166930305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,8192,0.027296000056796606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,7168,0.024703111913469102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,65536,0.2360124323103163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,8192,0.0485831101735433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,65536,0.11060355769263373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,8192,0.02794577678044637
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,6144,0.023588443795839947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,7168,0.04638133446375529
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,6144,0.046366221374935575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,5120,0.021760889225535925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,5120,0.042113777663972646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,7168,0.02528711160024007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,4096,0.020256888535287645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,3584,0.01331733332739936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,3584,0.019484443797005545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,5120,0.019854222734769184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,6144,0.022568888134426538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,4096,0.03786044319470724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,4096,0.01681866745154063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,3072,0.01202933324707879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,3584,0.037951111793518066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,3072,0.018789332773950364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,2560,0.010304888917340172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,2560,0.01797155539194743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,3072,0.037045333120558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,2048,0.009010666774378883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,2560,0.03548444310824076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,2048,0.017237333787812125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,1536,0.007718222008811102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,2048,0.033964445193608604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,1536,0.014639111028777229
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,1024,0.006126222097211414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,65536,0.1826311084959242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,1536,0.033775111039479576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,1024,0.013579555683665805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,768,0.00535733335547977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,1024,0.029066665305031672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,768,0.013697778185208639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,512,0.0037208890749348533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,768,0.029680000411139593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,512,0.013077333569526672
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,256,0.00328533330725299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,512,0.027943111128277246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,256,0.013156444662147097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,128,0.0030657777355776895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,256,0.02829155657026503
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,128,0.012291555603345236
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,64,0.0028702221396896574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,32,0.0028684443483750024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,128,0.025248888466093276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,64,0.011728000309732227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,32,0.011548444628715515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,16384,0.037457777394188776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,16384,0.06651200188530816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,16384,0.046524445215861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,65536,0.18635466363694933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,12288,0.03163822160826789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,12288,0.03557866811752319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,10240,0.029225776592890423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,12288,0.05849777989917331
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,65536,0.09832355711195205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,8192,0.025478222303920325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,8192,0.0468791127204895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,7168,0.02377066678471035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,10240,0.030622220701641504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,10240,0.05504622062047323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,7168,0.0440577765305837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,6144,0.02381244467364417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,8192,0.025444444682863023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,6144,0.04577955603599548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,5120,0.021341333786646526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,7168,0.023060444328520033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,4096,0.013391999734772576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,5120,0.04166933231883579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,65536,0.1700915495554606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,4096,0.020033778415785897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,3584,0.012159999873903064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,6144,0.020488889680968392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,4096,0.037584000163608126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,3584,0.018951111369662814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,5120,0.01831377711560991
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,3072,0.011037333144081963
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,3072,0.0185208883550432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,3584,0.036815110180113055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,2560,0.009736888938479954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,2560,0.017658667431937326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,3072,0.03628177775277032
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,2048,0.00847822262181176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,2560,0.035051554441452026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,2048,0.016739555531077914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,1536,0.007232888705200619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,2048,0.03460622164938185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,1536,0.014160000615649752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,1024,0.006025777922736273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,1536,0.030156443516413372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,1024,0.013516444298956128
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,768,0.004020444634887907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,1024,0.029470221863852605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,768,0.013730666703648038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,512,0.003671110918124517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,768,0.029412445094850328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,512,0.012880889077981314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,256,0.003286222202910317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,512,0.028280887338850234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,256,0.013209777573744455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,128,0.003030222323205736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,256,0.028198222319285076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,128,0.012282667060693106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,64,0.0028417776856157514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,32,0.0028595555987623003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,128,0.024960888756646052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,64,0.011796444654464722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,32,0.011600000162919363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,16384,0.03362755642996894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,16384,0.06439733505249023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,12288,0.030952887402640447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,65536,0.1641777753829956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,16384,0.039806223577923246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,12288,0.02920177910063002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,10240,0.026526222626368206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,12288,0.056926223966810435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,10240,0.054526223076714404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,8192,0.023711111810472276
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,7168,0.022259554929203455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,8192,0.046190222104390465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,10240,0.02689244515366024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,7168,0.044847998354170054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,65536,0.08553422159618801
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,8192,0.022699554761250813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,6144,0.021617778473430212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,6144,0.04240977764129639
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,5120,0.016252444850073922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,7168,0.020575111111005146
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,5120,0.02004888819323646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,5120,0.0414648883872562
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,6144,0.018373333745532565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,4096,0.014816888504558139
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,4096,0.019111111760139465
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,3584,0.013623110949993134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,3584,0.018897778458065458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,4096,0.03752177622583177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,3072,0.012426666915416718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,3584,0.03719288773006863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,3072,0.017806222041447956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,2560,0.009090666969617208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,3072,0.03611200054486593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,65536,0.1618702279196845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,2560,0.017120889491505094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,1536,0.006715555571847492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,2048,0.007691555553012424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,2560,0.03480444351832072
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,2048,0.015215999550289579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,2048,0.03355911042955186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,1536,0.03247111042340597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,1536,0.014087110757827759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,1024,0.005609777652555042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,1024,0.013408000270525614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,768,0.004371555729044808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,1024,0.028866665230857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,768,0.013610666824711693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,512,0.003649777836269803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,768,0.029467556211683486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,512,0.012840888566441007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,256,0.0033128888656695685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,512,0.02801777919133504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,256,0.013105777402718862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,128,0.002986666642957263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,256,0.02831022275818719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,128,0.012334222594896952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,64,0.003063999944263034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,128,0.02479733361138238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,32,0.002854222224818336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,64,0.01166133334239324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,32,0.011674666570292579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,65536,0.14348355929056802
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,16384,0.030932443009482488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,16384,0.06404178010092841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,12288,0.026733333865801494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,12288,0.05411555700831943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,16384,0.03371466530693902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,10240,0.02375288969940609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,12288,0.027105778455734253
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,10240,0.05492266681459215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,8192,0.021928000781271193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,8192,0.047192000680499606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,7168,0.0209840006298489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,10240,0.0236151119073232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,6144,0.016403555870056152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,7168,0.04557333389918009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,6144,0.020261334048377145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,8192,0.01979022224744161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,7168,0.0181048893266254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,5120,0.014412444498803882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,6144,0.041760888364579946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,5120,0.019170666734377544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,4096,0.012668444050682915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,3584,0.011922666596041786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,4096,0.018581334087583754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,5120,0.041189332803090416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,3584,0.017800889081425138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,4096,0.03808533483081394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,65536,0.07487110959159003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,3072,0.0110871113008923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,3584,0.03657955593532986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,65536,0.1450106700261434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,3072,0.01735111077626546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,2560,0.008223111430803934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,3072,0.03583200110329522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,2560,0.015266666809717814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,2048,0.007335111498832703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,2560,0.03457599878311157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,2048,0.014664888381958008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,1536,0.006507555643717448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,1536,0.03119022316402859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,2048,0.03257600135273404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,768,0.004029333177540037
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,1024,0.013426666458447775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,1536,0.01404355631934272
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,1024,0.0046880000995265115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,1024,0.028697778781255085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,768,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,512,0.003583111282851961
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,768,0.029234667619069416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,256,0.01313244468635983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,512,0.012840000291665396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,256,0.0032302221904198327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,512,0.027775999572541978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,128,0.0029893333299292457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,256,0.028078221612506445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,128,0.012345777617560493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,64,0.0028195555011431375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,128,0.02474755545457204
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,32,0.0028657778683635923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,64,0.011599111060301462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,32,0.011719110939237805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,16384,0.028932445579104956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,16384,0.06732177734375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,16384,0.027837332752015855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,12288,0.025166221790843542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,12288,0.022610666023360357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,65536,0.12021866109636094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,10240,0.022572444544898138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,12288,0.05427111188570658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,10240,0.05448266532686022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,8192,0.02120977805720435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,8192,0.049039999643961586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,7168,0.01531644496652815
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,10240,0.01979200045267741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,7168,0.020394666327370536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,7168,0.04348799917432997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,6144,0.013788444300492605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,6144,0.019440889358520508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,8192,0.01676177812947167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,65536,0.06779022349251641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,5120,0.012864888542228274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,5120,0.018413333429230582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,6144,0.04105244411362542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,4096,0.011429333024554782
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,4096,0.01780533293883006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,5120,0.041013333532545306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,3584,0.010407110883129967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,4096,0.03843644592497084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,3584,0.01664444473054674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,3584,0.03592533204290602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,3072,0.010209778116808997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,65536,0.16360355748070612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,3072,0.016546666622161865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,2560,0.00777599960565567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,3072,0.03538844320509169
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,2560,0.015014222926563688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,2048,0.006879111131032308
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,2560,0.03241422110133701
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,2048,0.014488889111412896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,1536,0.013847110999955071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,1536,0.005943111247486538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,2048,0.0306328899330563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,1024,0.004485333131419288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,1536,0.02987288766437107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,1024,0.013316444224781461
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,768,0.01367022262679206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,768,0.003823111040724648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,1024,0.028749333487616643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,512,0.0034826666944556763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,768,0.02948533495267232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,512,0.012825777961148156
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,256,0.0031724443866146933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,512,0.0277173337009218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,256,0.013073777986897362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,128,0.012219555675983429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,256,0.02823644545343187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,128,0.0029333333174387612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,64,0.0027866665687825945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,128,0.024895111719767254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,32,0.0028239999794297745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,64,0.011609777808189392
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,32,0.01146044416560067
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,16384,0.026945778065257605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,16384,0.060995552274915904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,65536,0.09652533133824666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,12288,0.022036444809701707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,10240,0.016511110795868766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,12288,0.052505777941809766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,10240,0.020936000678274367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,8192,0.014049778381983439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,12288,0.01867555578549703
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,10240,0.05172355638609993
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,8192,0.01947022146648831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,16384,0.022712000542216833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,7168,0.013372444444232516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,8192,0.04768622252676222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,7168,0.019886222150590684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,6144,0.012325333224402534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,7168,0.043376889493730336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,6144,0.01906577746073405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,5120,0.011041777829329172
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,6144,0.04060711132155524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,5120,0.017887110511461895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,4096,0.010063110954231685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,5120,0.040579554107454084
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,4096,0.016938666502634685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,65536,0.06222044759326511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,3584,0.009655999640623728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,3584,0.01568800045384301
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,3072,0.008992888861232335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,4096,0.03767111235194736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,3584,0.03587466809484694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,3072,0.01497244503762987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,2560,0.014642667439248828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,2560,0.007177777588367462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,3072,0.032525334093305804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,2048,0.006266666783226862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,2560,0.031192888816197712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,2048,0.014236445228258768
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,1536,0.005174222091833751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,1024,0.004127999974621667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,2048,0.030471109681659277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,1536,0.013775111072593264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,65536,0.14200622505611843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,1536,0.030093332131703694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,768,0.0038008888562520346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,1024,0.013181333740552267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,1024,0.02971200148264567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,768,0.013663999736309052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,512,0.00350222239891688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,512,0.027758222487237718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,768,0.0288426677385966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,512,0.01296622223324246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,256,0.003198222153716617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,256,0.013056000073750814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,128,0.0029528888149393927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,256,0.02798488736152649
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,128,0.012103111379676394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,64,0.002801777794957161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,128,0.024840889705551997
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,32,0.0028071111689011254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,64,0.011550222006109027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,32,0.011414222419261932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,65536,0.07453866799672444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,16384,0.023616888456874426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,12288,0.014352889524565803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,16384,0.05822844637764824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,12288,0.021572444174024794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,10240,0.012820444173283048
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,12288,0.050254222419526845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,10240,0.02017866737312741
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,16384,0.017027555240525138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,8192,0.01144444445768992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,10240,0.0482666658030616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,8192,0.01921066641807556
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,7168,0.01072355525361167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,8192,0.0428746673795912
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,7168,0.018746667438083224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,6144,0.009777777724795872
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,65536,0.050936887661616005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,7168,0.04440444376733568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,6144,0.016946666770511203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,5120,0.009339555270142024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,6144,0.039048890272776283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,4096,0.008786666724416945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,5120,0.015955555770132277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,4096,0.015441776977645027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,5120,0.03469511204295688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,3584,0.008626666333940294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,3584,0.015094222293959724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,65536,0.12331199645996094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,4096,0.03281155559751723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,3072,0.00790666706032223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,3584,0.0321742229991489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,3072,0.014911999305089315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,2560,0.006522666662931442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,3072,0.0325528879960378
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,2560,0.014523555835088095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,2048,0.0057795556883017225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,2560,0.030398223135206435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,2048,0.029692444536421034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,2048,0.014224888549910652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,65536,0.05021244287490845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,1536,0.005043555465009478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,1536,0.013817778064144982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,1024,0.0041653331783082755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,1536,0.028900444507598877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,1024,0.013325333595275879
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,768,0.0289066665702396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,768,0.003826666623353958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,1024,0.028206222587161597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,768,0.013538666897349887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,128,0.0029271110478374693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,512,0.003491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,512,0.012998222476906247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,256,0.003175111073586676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,512,0.02749600013097127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,256,0.01313244468635983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,256,0.027863999207814533
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,64,0.0031537777847713898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,128,0.012215110990736218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,32,0.003109333415826162
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,128,0.02475999957985348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,64,0.011713777979214987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,32,0.011517333487669626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,16384,0.014494222071435718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,16384,0.022590221630202398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,12288,0.012528888881206512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,16384,0.053660445743136935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,12288,0.020534222324689228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,10240,0.011438222395049201
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,10240,0.01978755493958791
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,12288,0.04755199948946635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,8192,0.010429333481523724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,10240,0.045575112104415894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,8192,0.01814311080508762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,7168,0.00962222201956643
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,7168,0.017151999804708693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,8192,0.04082577758365207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,6144,0.008905777500735389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,7168,0.04221422142452664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,6144,0.016327111257447135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,5120,0.008711111214425828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,5120,0.015764445066452026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,65536,0.04567022124926249
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,6144,0.03593066665861342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,4096,0.00756533361143536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,5120,0.03350222110748291
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,4096,0.015375999940766228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,65536,0.11364089118109809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,3072,0.007245333658324347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,3584,0.007502222226725684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,3584,0.031566222508748375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,4096,0.03193511234389411
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,3584,0.014915555715560913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,65536,0.03953511185116238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,2560,0.006490666833188798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,3072,0.014833778142929077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,2560,0.014560000763999091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,3072,0.03201866812176175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,2048,0.005475555443101459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,2560,0.029586666160159644
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,2048,0.01402666668097178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,1536,0.004822222308980094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,2048,0.029192888074451025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,1024,0.013262222210566202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,1536,0.013746666411558786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,1024,0.028060442871517603
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,1024,0.004108444270160464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,1536,0.028805332051383123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,768,0.003801777958869934
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,768,0.013584888643688627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,512,0.003499555504984326
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,768,0.028708444701300725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,512,0.012679110798570844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,256,0.00328533330725299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,512,0.02786311176088121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,256,0.013181333740552267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,128,0.003079111170437601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,256,0.027833776341544256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,128,0.01222400036123064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,64,0.0029208889851967492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,128,0.024707555770874023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,32,0.0029288888391521242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,64,0.011628444823953839
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,32,0.011509333219793109
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,16384,0.011747555600272285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,16384,0.022421333524915908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,12288,0.01014666673209932
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,16384,0.048413332965638906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,12288,0.01994044416480594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,10240,0.009579555855857002
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,12288,0.04346133271853129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,10240,0.01808088852299584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,8192,0.008970666262838576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,65536,0.04539822207556831
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,10240,0.041365332073635526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,7168,0.008309333688682979
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,8192,0.01718755563100179
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,6144,0.0074417781498697065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,8192,0.03673244516054789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,7168,0.016514667206340365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,6144,0.016329778565300834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,7168,0.03744355506367154
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,65536,0.10071644518110488
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,5120,0.007734222544564142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,6144,0.033025778002209134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,5120,0.01573333309756385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,4096,0.006680000159475539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,5120,0.032307555278142296
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,65536,0.028741333219740126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,4096,0.01533777846230401
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,3584,0.00758133331934611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,4096,0.030961778428819444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,3584,0.015066666735543145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,3072,0.006836444553401735
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,3584,0.030224889516830444
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,2560,0.006287999865081575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,3072,0.014614222778214348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,2560,0.014527999692493014
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,3072,0.030995554394192163
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,2048,0.005648889061477449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,2560,0.02928000026279026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,2048,0.014260444376203748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,1536,0.004924444274769889
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,2048,0.028704888290829126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,1536,0.01370044466522005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,1024,0.004260444392760594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,1536,0.028466665082507666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,1024,0.013223999904261695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,768,0.003907555507289039
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,1024,0.027875555886162653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,768,0.01366666704416275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,512,0.003568888952334722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,768,0.028447998894585505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,512,0.01275022245115704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,256,0.0032808888289663526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,512,0.027593778239356145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,256,0.013168888787428537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,128,0.003002666764789157
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,256,0.02776977750990126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,64,0.0028355556229750314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,128,0.01219911128282547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,128,0.024677332904603746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,64,0.011508444117175208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,16384,0.00885511106914944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,32,0.0028311111446883944
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,16384,0.021537777450349595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,32,0.011617778076065911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,65536,0.043819556633631386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,12288,0.008096000386608971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,65536,0.09411466783947414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,16384,0.044857776827282377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,12288,0.018997333116001554
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,10240,0.007652444144090016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,12288,0.03972533345222473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,65536,0.017584888471497428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,10240,0.017684444785118103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,8192,0.006990222467316522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,10240,0.037816888756222196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,8192,0.017147555947303772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,7168,0.006743999818960826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,7168,0.016716443830066256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,8192,0.03503377901183234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,5120,0.006996444529957241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,6144,0.006671999891599019
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,7168,0.033709334002600774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,6144,0.016210666961140104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,6144,0.0324888891643948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,5120,0.015810666812790763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,4096,0.006616000086069107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,5120,0.03225955698225234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,4096,0.015278221832381355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,3584,0.007040888898902469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,4096,0.03082044588194953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,3584,0.0151235560576121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,3072,0.006744888921578725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,3584,0.030441777573691473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,2560,0.014364444547229342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,3072,0.014648000399271647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,2560,0.006127111199829314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,3072,0.02982577681541443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,2048,0.005481777919663324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,2560,0.02908266584078471
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,2048,0.014167111780908374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,1536,0.004840888910823398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,2048,0.02866666515668233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,1536,0.013599110974205865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,1024,0.004139555411206352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,1536,0.02832888894610935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,1024,0.013211555778980255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,768,0.0037848887344201407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,1024,0.027969777584075928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,768,0.013429332938459186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,512,0.0034622223012977173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,768,0.028247998820410833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,512,0.012702222499582501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,256,0.003192000091075897
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,512,0.027254222167862788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,256,0.013023999830087027
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,128,0.0029235554652081597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,256,0.0277786652247111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,128,0.012103111379676394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,64,0.0028275555620590844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,128,0.02463111115826501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,32,0.0028248888750871024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,64,0.011559111376603445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,32,0.011376889215575324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,65536,0.012652444342772165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,65536,0.043416887521743774
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,16384,0.006818666640255187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,12288,0.006696888970004187
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,65536,0.09382488992479111
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,16384,0.020488000578350492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,16384,0.04359289010365804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,12288,0.01870666609870063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,10240,0.006647999915811751
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,12288,0.03944889042112563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,10240,0.017492444978819955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,8192,0.006719999843173557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,10240,0.03770222266515096
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,8192,0.01681600014368693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,7168,0.006522666662931442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,8192,0.03481066558096144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,7168,0.016368889146380954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,6144,0.0064177778032090925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,7168,0.03379288978046841
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,6144,0.015776000089115567
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,5120,0.006702222343948152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,6144,0.032419555717044406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,5120,0.015557333827018738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,4096,0.006314666734801398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,3584,0.029987557066811457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,5120,0.031882666879230075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,4096,0.015142222245534262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,3584,0.006919999917348226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,3072,0.02959377898110284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,4096,0.030633777379989624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,3584,0.014652444256676568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,3072,0.00646311127477222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,3072,0.014352889524565803
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,2560,0.0058408888677755995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,2560,0.014073777529928418
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,2048,0.005214222189452913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,2560,0.029725333054860432
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,2048,0.013745778136783175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,1536,0.004597333156400257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,1024,0.013073777986897362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,2048,0.02858755654758877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,1536,0.013367999758985309
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,1024,0.003925333420435588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,1536,0.02854577700297038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,768,0.028155555327733357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,512,0.012610666453838348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,768,0.0037573331760035623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,1024,0.028107557031843398
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,512,0.003478222423129611
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,768,0.013454222016864352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,256,0.003128000017669466
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,512,0.02714666724205017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,256,0.012851555314328937
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,128,0.0029493332323100832
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,256,0.027678221464157104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,128,0.012158222496509552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,64,0.0027964444210131965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,128,0.024538666009902954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,32,0.0027635554886526535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,64,0.01146577795346578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,32,0.011506666739781698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,65536,0.00979466653532452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,16384,0.00646311127477222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,65536,0.04275022281540764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,12288,0.006405333264006509
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,16384,0.020269332660569083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,10240,0.006887110984987683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,12288,0.018378666705555387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,8192,0.006567111031876669
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,10240,0.017364444004164804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,7168,0.006384000182151794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,8192,0.01660711069901784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,6144,0.00646666685740153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,7168,0.01615111033121745
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,5120,0.00673333348499404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,6144,0.015631111131774057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,4096,0.006249777972698212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,5120,0.015180443723996481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,3072,0.01408088869518704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,3584,0.006847111301289664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,4096,0.014846222268210517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,3072,0.006485333459244833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,3584,0.014504000544548035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,2560,0.005684444473849402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,2048,0.005118222286303838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,2560,0.01390044391155243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,1536,0.004582222137186262
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,2048,0.013872000078360239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,1024,0.0038684445122877755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,1536,0.013314666847387949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,768,0.0035866668654812705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,1024,0.012882666455374824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,512,0.003294222263826264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,768,0.013139555851618448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,256,0.003105777833196852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,512,0.012832000023788877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,64,0.011519110865063138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,128,0.002946666752298673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,256,0.012991111311647626
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,64,0.002814222127199173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,128,0.012072000238630505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,16384,0.00666844430896971
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,32,0.002777777819169892
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,65536,0.008721777962313758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,32,0.01146399974822998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,65536,0.04191999965243869
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,12288,0.006508444332414203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,16384,0.020417778028382193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,12288,0.0185253338681327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,10240,0.006604444649484422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,8192,0.006522666662931442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,7168,0.01633155511485206
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,10240,0.01754844519827101
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,7168,0.0063697778516345555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,8192,0.01678400072786543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,6144,0.006321777900060018
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,5120,0.006659555352396435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,6144,0.015857777661747403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,4096,0.006267555471923616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,5120,0.015464888678656684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,4096,0.0147724449634552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,3584,0.006711111300521427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,3072,0.0064311110311084324
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,3584,0.014560000763999091
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,2560,0.005722666780153911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,3072,0.01418133411142561
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,2048,0.005179555465777715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,2560,0.01404800017674764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,1536,0.004519999855094486
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,2048,0.01368977791733212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,1024,0.0038968887594011095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,1536,0.013388444152143268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,1024,0.013025778035322825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,768,0.0036311112344264984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,512,0.0032835555159383347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,768,0.013432888521088494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,256,0.002959111084540685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,512,0.012633778154850006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,128,0.0028515555378463534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,256,0.012977778083748288
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,64,0.002712888850106133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,128,0.012229333321253458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,32,0.0027840000887711844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,64,0.011437333292431302
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,32,0.01144266708029641
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,12288,0.2765608893500434
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,10240,0.23293511072794595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,10240,0.47185956107245547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,16384,0.3636524412367079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,12288,0.5652524630228678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,8192,0.21633244885338676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,12288,0.3467511071099176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,16384,0.7439902093675402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,16384,0.41501421398586696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,7168,0.1676257716284858
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,8192,0.3783893320295546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,7168,0.3290453222062853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,10240,0.29560089111328125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,8192,0.2447813351949056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,6144,0.14612800545162624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,5120,0.14389333460066053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,6144,0.3255564371744792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,5120,0.2358542283376058
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,4096,0.10232977734671699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,7168,0.21738578213585746
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,3584,0.10607822073830499
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,4096,0.19010844495561388
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,6144,0.18794666396247017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,3584,0.16639822059207493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,3072,0.08153422011269464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,2560,0.07746311028798421
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,3072,0.1527235507965088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,2560,0.12042666806115045
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,5120,0.16154400507609049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,4096,0.12973866197797987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,2048,0.09623555342356364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,2048,0.06001777781380547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,1536,0.05061422122849358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,1024,0.0401777790652381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,1536,0.07388089100519817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,1024,0.050127999650107495
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,768,0.03720622261365255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,3584,0.10971289210849339
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,768,0.038908445172839694
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,512,0.032891554964913264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,512,0.02683288852373759
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,256,0.015141333142916361
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,256,0.028282665544086035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,3072,0.10329333278867935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,128,0.010336000058386061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,128,0.025507556067572698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,128,0.03090311090151469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,64,0.006375999914275275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,64,0.023000000251664057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,32,0.004562666846646203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,32,0.02239466706911723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,2560,0.09380444553163315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,2048,0.08223555485407512
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,1536,0.06589511368009779
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,1024,0.0532604455947876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,768,0.04850488901138306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,512,0.041999111572901406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,256,0.035084442959891424
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,16384,0.10348088873757257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,16384,0.18419022030300566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,12288,0.08224533663855658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,16384,0.13200178411271837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,12288,0.10732266638014053
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,12288,0.1398871077431573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,10240,0.07079911231994629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,10240,0.08811910947163899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,10240,0.11619910928938125
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,8192,0.09414311250050862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,8192,0.059248891141679555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,8192,0.07574133078257243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,7168,0.05429511268933614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,7168,0.0745404428905911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,65536,0.3712008794148763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,6144,0.048728889889187284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,7168,0.0830577810605367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,6144,0.06406044297748141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,6144,0.07191999753316243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,5120,0.04771733283996582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,5120,0.060066666867997914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,5120,0.0583546691470676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,4096,0.03802311089303758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,4096,0.05197333296140035
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,3584,0.035707556539111666
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,4096,0.04888000090916952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,3584,0.05079022381040785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,3072,0.037598222494125366
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,3584,0.04309244288338555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,3072,0.0328746669822269
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,3072,0.047190222475263804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,2560,0.030322667625215318
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,2048,0.026733333865801494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,2560,0.04523644513554043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,2560,0.03176800078815884
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,1536,0.024118221468395654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,2048,0.04064622190263536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,1536,0.037853333685133196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,1024,0.020922667450375024
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,2048,0.02581866747803158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,768,0.011678222152921887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,1024,0.034167998366885714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,1536,0.020436444216304354
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,512,0.008827555510732863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,768,0.02108000053299798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,768,0.03372800019052293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,256,0.005960888746711943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,1024,0.014544000228246054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,512,0.018999111321237352
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,512,0.032023111979166664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,256,0.014215999179416232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,128,0.00360000009338061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,256,0.02605955633852217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,128,0.013420444395807056
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,64,0.0032044444233179092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,32,0.0032151111712058387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,128,0.024656888511445787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,64,0.013444444371594323
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,32,0.013302221894264221
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,16384,0.08312177658081055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,65536,0.42818133036295575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,16384,0.11261510848999023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,16384,0.14488977856106228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,12288,0.07248977820078532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,12288,0.08933155404196845
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,12288,0.11059911383522882
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,10240,0.05785510937372843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,10240,0.08110311296251085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,10240,0.08902221918106079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,65536,0.7397306760152181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,8192,0.04940711127387153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,8192,0.0756782227092319
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,8192,0.06889333327611287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,7168,0.05047022302945455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,7168,0.06395555867089166
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,65536,0.2839591238233778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,6144,0.041165331999460854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,65536,0.36500710911220974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,7168,0.06344622373580933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,6144,0.06374310784869723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,6144,0.059416890144348145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,5120,0.03819200065400865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,5120,0.0543431109852261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,4096,0.03351199958059523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,5120,0.04542577928966946
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,4096,0.03723911113209195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,4096,0.04943199952443441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,3584,0.031311111317740545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,3584,0.046663112110561795
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,3072,0.02868088748719957
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,2560,0.02609155575434367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,3584,0.03257955445183648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,3072,0.04478222131729126
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,2048,0.02348355617788103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,2560,0.041722665230433144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,1536,0.022239110536045496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,3072,0.028588443994522095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,2048,0.03771911064783732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,1024,0.011691555380821228
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,1536,0.035994665490256414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,1024,0.019923556182119582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,2048,0.02032177812523312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,768,0.009947555760542551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,1536,0.015864888827006023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,2560,0.024898666474554274
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,1024,0.03288177649180094
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,768,0.01928711103068458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,512,0.007609777980380588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,768,0.03298844562636481
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,256,0.013987556099891663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,512,0.016163556112183463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,256,0.004433777597215441
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,512,0.03122399912940131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,128,0.0035955554081334006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,256,0.02570933269129859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,128,0.013396444420019785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,64,0.0033271111961868075
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,128,0.02440711193614536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,32,0.0033173333439562055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,64,0.013101333545313941
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,32,0.013064888616402945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,16384,0.07228710916307238
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,16384,0.08941955698861016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,12288,0.0893591112560696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,12288,0.05817955732345581
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,16384,0.11642489168379043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,10240,0.05123111274507311
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,12288,0.07475466860665216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,10240,0.07450577947828504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,8192,0.044023109806908496
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,10240,0.06798933611975776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,8192,0.0639937784936693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,8192,0.06017333269119263
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,65536,0.553034676445855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,7168,0.040891554620530866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,7168,0.05704889032575819
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,6144,0.04109333289994134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,7168,0.05316355493333605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,6144,0.046647999021742076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,6144,0.052022222015592784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,5120,0.03916888766818576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,5120,0.03405333227581448
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,65536,0.23743645350138345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,5120,0.04802311129040188
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,4096,0.03048355711830987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,4096,0.03175110949410333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,4096,0.04431644413206312
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,3072,0.026241777671708003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,3584,0.02935911218325297
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,3584,0.04366222355100843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,3072,0.04112622141838074
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,3072,0.02478844424088796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,2560,0.024876443876160517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,3584,0.02906044324239095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,2560,0.03853955533769395
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,2048,0.02239466706911723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,1536,0.013896889156765409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,2048,0.03664533297220866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,1536,0.021287999219364587
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,1024,0.010425777898894416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,2560,0.0210826661851671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,1536,0.03489777776930068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,1024,0.018979556030697294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,2048,0.01739377776781718
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,768,0.008543999658690559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,1024,0.03270399901602004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,768,0.018730666902330186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,512,0.006737777756320105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,65536,0.30632978015475804
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,768,0.0324524442354838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,256,0.003727999826272329
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,512,0.015471110741297403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,256,0.013768889009952545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,128,0.003379555626047982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,512,0.03106044398413764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,256,0.025751110580232408
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,128,0.013106666505336761
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,64,0.003072000005178981
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,128,0.024393777052561443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,64,0.012999110751681857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,32,0.003069333318206999
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,32,0.013088888592190213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,16384,0.06013066901101006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,16384,0.07985511091020372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,12288,0.07041422526041667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,65536,0.4561324649386936
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,12288,0.049667557080586754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,16384,0.09222933318879868
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,12288,0.06717689169777764
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,10240,0.04822577701674568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,10240,0.06328177452087402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,8192,0.037668443388409086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,10240,0.05891022417280409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,8192,0.05357244279649523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,8192,0.047842668162451855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,7168,0.034958223501841225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,65536,0.19362221823798287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,7168,0.05240799983342489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,6144,0.032207998964521624
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,7168,0.042388445801205106
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,6144,0.03720000055101182
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,6144,0.04758755697144402
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,5120,0.029306666718588933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,5120,0.04443555408053928
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,5120,0.03131377696990967
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,4096,0.02548888822396596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,3584,0.02442933287885454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,3584,0.039798223310046725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,3072,0.02249777813752492
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,4096,0.025655110677083332
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,4096,0.041119111908806696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,3072,0.03757244348526001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,2560,0.02167822255028619
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,3072,0.02012355625629425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,3584,0.022787556052207947
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,2048,0.019647111495335896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,2560,0.03659466571278042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,2048,0.03413688805368211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,1536,0.018924444913864136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,1536,0.01163999984661738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,2560,0.01735466718673706
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,1024,0.008640000389681922
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,2048,0.014257777068350049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,1536,0.03357866737577651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,1024,0.016883555385801528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,768,0.007439999944633908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,512,0.005752888818581899
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,768,0.015999111864301894
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,768,0.03155733479393853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,256,0.003587555554178026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,1024,0.03205600049760606
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,512,0.0140666663646698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,128,0.003277333246337043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,256,0.012543111211723752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,512,0.027686221732033625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,128,0.01178311142656538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,256,0.025455110602908667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,64,0.003000000077817175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,32,0.003026666740576426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,128,0.024420445164044697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,64,0.011776000261306763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,65536,0.25695021947224933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,32,0.011632888681358762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,16384,0.05500088797675239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,16384,0.07714133130179511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,65536,0.37282755639817977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,12288,0.04541777902179294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,16384,0.0875013338194953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,12288,0.06620089213053386
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,12288,0.06989600261052449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,10240,0.04028444488843282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,10240,0.058258665932549365
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,8192,0.035473777188195124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,8192,0.0462702214717865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,10240,0.0550799998972151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,8192,0.052213334374957614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,7168,0.03304444419013129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,7168,0.04875377813975016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,6144,0.030069334639443293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,7168,0.041441778341929116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,6144,0.04632000128428141
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,65536,0.17694489161173502
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,6144,0.035840888818105064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,5120,0.027432888746261597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,5120,0.030583110120561387
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,4096,0.023966223001480103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,4096,0.04002133342954848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,3584,0.023152000374264185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,4096,0.02312799957063463
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,5120,0.04633688926696777
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,3584,0.03904799951447381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,3072,0.021359110871950786
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,2560,0.02050044470363193
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,3072,0.03643911083539327
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,2560,0.035884443256590105
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,2048,0.013215111361609565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,3584,0.020492444435755413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,2048,0.018868444694413077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,3072,0.018405333161354065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,1536,0.010787555740939246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,1536,0.018100443813535903
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,1024,0.008151111503442129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,2048,0.03363022208213806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,2560,0.015971556305885315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,1536,0.033139556646347046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,1024,0.016139555308553908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,768,0.00702044450574451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,1024,0.030976000759336684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,768,0.014114666316244336
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,512,0.005666666560702854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,768,0.029399110211266413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,512,0.01371022231049008
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,65536,0.22684621810913086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,256,0.0035546666218174826
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,512,0.02752977775202857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,256,0.012527999778588613
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,128,0.0032319999817344877
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,256,0.025272889269722834
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,64,0.0029760001020299066
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,128,0.011725333001878528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,32,0.003007111036115222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,128,0.0243484444088406
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,64,0.01165066659450531
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,32,0.01185955521133211
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,16384,0.05198488963974846
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,16384,0.07453599903318617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,12288,0.057118223773108594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,16384,0.07502310805850558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,65536,0.3205804559919569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,12288,0.042596442831887134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,10240,0.037775112522972956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,12288,0.06409244404898749
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,10240,0.0586951110098097
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,8192,0.033407999409569636
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,10240,0.04833689000871447
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,8192,0.05005866620275709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,8192,0.03847644395298428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,7168,0.03121688961982727
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,7168,0.03494933247566223
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,6144,0.028901331954532202
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,7168,0.04776622189415825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,5120,0.025628444221284654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,6144,0.04468533396720886
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,5120,0.04224710994296604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,6144,0.03029511041111416
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,4096,0.02307111190425025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,4096,0.038776000340779625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,3584,0.02198933396074507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,5120,0.025880000657505457
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,65536,0.1615013281504313
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,3584,0.03829777903027005
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,3072,0.0207324441936281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,4096,0.021389333738221064
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,2560,0.014749333262443542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,3584,0.019052444232834708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,3072,0.036002665758132935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,2560,0.019917334119478863
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,3072,0.017087111870447796
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,2048,0.012467555701732635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,2560,0.03544266687499152
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,2048,0.018345778187115986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,1536,0.01016622202263938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,1536,0.017796445224020217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,2048,0.03382844395107693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,1024,0.007711999946170383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,1024,0.014557333456145393
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,1536,0.03338044550683763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,768,0.00638133328821924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,1024,0.030689779255125258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,512,0.005209777918126848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,768,0.014151111245155334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,512,0.013629333012633853
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,65536,0.20981155501471627
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,256,0.0034808889031410217
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,768,0.02866666515668233
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,512,0.027581334114074707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,256,0.012630222572220696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,128,0.0032035555276605818
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,64,0.0029404444826973807
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,128,0.011891555454995898
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,256,0.02544266647762722
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,64,0.01168444421556261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,128,0.02416533397303687
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,32,0.0029475556479560006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,32,0.011640888949235281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,16384,0.04571466644605001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,16384,0.07041777504814996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,65536,0.27949778238932294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,12288,0.037997331884172224
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,16384,0.058916442924075656
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,12288,0.06157777706782023
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,10240,0.03423110975159539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,10240,0.03846044341723124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,12288,0.046153777175479464
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,10240,0.05397689011361864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,8192,0.03055822186999851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,8192,0.03144177794456482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,8192,0.050107555256949536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,6144,0.02622311148378584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,7168,0.028408000866572063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,7168,0.04649688800175985
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,7168,0.028225776222017076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,6144,0.043121778302722506
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,5120,0.023801777097913954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,6144,0.024711999628278945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,4096,0.021673777037196692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,5120,0.041136887338426374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,4096,0.03779110974735684
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,3584,0.021087111698256597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,4096,0.017423111531469557
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,5120,0.02110044492615594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,65536,0.1357813278834025
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,3072,0.014153778553009033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,3584,0.03704799877272712
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,3072,0.019719999697473314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,3584,0.01571111049917009
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,3072,0.03548977772394816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,2560,0.013912000589900546
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,2560,0.019112888309690688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,2048,0.010411555568377176
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,2560,0.03493155704604255
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,2048,0.01750666730933719
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,1536,0.008752889103359645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,1536,0.017133333616786532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,2048,0.03291733397377862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,1024,0.006883555402358373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,1024,0.02991644541422526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,1536,0.0325582226117452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,1024,0.013561777770519257
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,768,0.005992888990375731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,512,0.00453599997692638
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,512,0.01349955548842748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,768,0.013987556099891663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,65536,0.17951644791497123
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,256,0.0033928888539473214
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,768,0.02956177790959676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,512,0.027500443988376196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,128,0.003149333306484752
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,256,0.01241422196229299
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,128,0.024012444747818842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,256,0.02532177832391527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,64,0.0028897778441508612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,128,0.011632888681358762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,32,0.0029288888391521242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,64,0.011669333610269757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,32,0.011630222201347351
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,16384,0.03994044330384996
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,16384,0.06622666782803006
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,12288,0.03936444388495551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,16384,0.050252444214291044
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,12288,0.03361599975162082
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,12288,0.06044266621271769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,10240,0.03071199854214986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,10240,0.05132977830039131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,8192,0.028093334701326158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,10240,0.03370044297642178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,65536,0.11022488938437568
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,7168,0.025105777713987563
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,8192,0.0273626661962933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,7168,0.044178667995664805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,65536,0.24389065636528862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,6144,0.023227555884255305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,8192,0.04637688729498121
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,7168,0.025161777933438618
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,6144,0.041639112763934664
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,5120,0.022038223014937505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,5120,0.03932711150911119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,4096,0.020599999361568026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,6144,0.022628444764349196
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,3584,0.01332711097266939
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,4096,0.0369484457704756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,65536,0.18365155325995552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,5120,0.019952888290087383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,4096,0.016840888394249808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,3584,0.019780443774329293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,3072,0.011752000285519494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,3072,0.018546667363908555
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,3584,0.03630933496687148
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,2560,0.010568888651000129
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,2560,0.018180444836616516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,3072,0.0357582237985399
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,2048,0.009102221992280748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,1536,0.00776533368561003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,2560,0.03445422318246629
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,2048,0.017239999439981248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,2048,0.03255911005867852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,1536,0.016170667277442086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,1024,0.006268444574541516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,1536,0.03221866819593642
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,1024,0.013382222089502545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,768,0.005062222066852781
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,1024,0.027760000692473516
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,768,0.014095111025704278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,512,0.0037493333220481873
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,768,0.028037332826190527
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,512,0.013427555561065674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,256,0.0032808888289663526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,128,0.0030782222747802734
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,512,0.027428444888856676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,256,0.012514666550689273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,256,0.025349333882331848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,64,0.0028640000770489373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,128,0.011589333415031433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,128,0.024001777172088623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,64,0.011552000211344825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,32,0.0029324444217814338
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,32,0.01145333300034205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,16384,0.035742223262786865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,16384,0.06925955745908949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,16384,0.04487377736303541
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,12288,0.030870222383075293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,12288,0.05976444482803345
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,65536,0.18648355536990693
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,12288,0.03567555546760559
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,10240,0.030270222160551283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,8192,0.02555999987655216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,10240,0.028443555037180584
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,10240,0.05097244514359368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,7168,0.023412444525294836
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,8192,0.044999112685521446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,65536,0.09821689128875732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,7168,0.04306933283805847
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,6144,0.022523555490705702
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,8192,0.02881066666709052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,5120,0.021399999658266704
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,6144,0.04094311263826158
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,5120,0.03898488812976413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,4096,0.013312000367376538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,7168,0.023141332798533972
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,4096,0.019723556107944913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,3584,0.012146666646003723
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,5120,0.018041777941915724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,6144,0.020766221814685397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,4096,0.03688622183269925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,3584,0.019123555885420907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,3072,0.010988444089889526
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,3584,0.03574133250448439
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,3072,0.03442222211096022
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,3072,0.01822933389080895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,2560,0.009766221874290043
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,2560,0.01787911189926995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,2048,0.008257778154479133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,2560,0.03401422169473436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,2048,0.016378665963808697
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,1536,0.007313778003056844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,1536,0.014507555299335055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,2048,0.0322088897228241
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,1024,0.005931555396980709
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,768,0.004660444541109933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,1536,0.02997066577275594
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,1024,0.01352266636159685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,1024,0.02738133403990004
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,512,0.0036924444139003754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,768,0.013773333695199756
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,65536,0.17183732986450195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,512,0.013339555925793119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,256,0.012320888539155325
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,768,0.027837332752015855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,256,0.0032960000551409195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,512,0.027609777119424608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,128,0.003056888985964987
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,256,0.02517333295610216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,128,0.011549333731333414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,64,0.002893333426780171
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,128,0.02400622268517812
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,32,0.0028657778683635923
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,64,0.011482666763994427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,32,0.011485333244005838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,16384,0.03352444370587667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,16384,0.06350844436221652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,12288,0.031166222360399034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,16384,0.040017777019076876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,12288,0.029203553994496662
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,12288,0.0547982222504086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,10240,0.02653955585426754
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,65536,0.16311644183264837
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,10240,0.04876533481809828
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,8192,0.025052444802390203
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,8192,0.0466595557000902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,65536,0.0851973361439175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,7168,0.020624889267815482
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,7168,0.022328888376553852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,10240,0.027218666341569688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,7168,0.042187554968727954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,6144,0.020997333857748244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,5120,0.0162782222032547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,5120,0.02028888960679372
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,8192,0.022346667117542688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,6144,0.040785776244269475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,4096,0.014829332629839579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,5120,0.0403386652469635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,6144,0.01827555563714769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,4096,0.01910044418440925
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,4096,0.036453333165910505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,3584,0.013405332962671915
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,3584,0.018803555104467604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,3072,0.012253333297040729
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,3584,0.035460445615980364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,3072,0.017831999394628737
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,2560,0.00905955582857132
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,3072,0.03427733315361871
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,2560,0.017453332742055256
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,2048,0.007735999921957652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,1536,0.006670222100284364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,2560,0.03378222386042277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,2048,0.0143857780430052
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,2048,0.031896889209747314
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,65536,0.15787022643619114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,1024,0.013263999587959714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,1024,0.005609777652555042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,1536,0.014330666926172046
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,1536,0.02959733208020528
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,768,0.004209777961174647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,1024,0.027356444133652583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,512,0.013529777526855469
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,768,0.013768889009952545
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,512,0.0036764442920684814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,256,0.012272888587580787
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,768,0.027660444378852844
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,256,0.0032995556377702286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,512,0.027121777335802715
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,64,0.0028515555378463534
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,128,0.003050666716363695
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,256,0.025489777326583862
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,128,0.011567999919255575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,128,0.023834667272037927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,32,0.0028737777223189673
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,64,0.011511111425028907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,32,0.011511111425028907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,16384,0.030877331892649334
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,65536,0.14314756128523085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,16384,0.061920000447167285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,12288,0.026351110802756414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,12288,0.05412888858053419
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,10240,0.023967110448413428
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,12288,0.026758222116364375
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,10240,0.04756089051564535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,8192,0.01973155637582143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,8192,0.022395556171735127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,16384,0.03422666589419047
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,8192,0.04422755704985725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,7168,0.018114666144053142
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,7168,0.021362667282422382
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,10240,0.023336889015303716
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,7168,0.0425893333223131
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,6144,0.01629155543115404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,6144,0.020080000162124634
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,5120,0.014508444401952954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,6144,0.040928890307744346
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,5120,0.019328888919618394
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,5120,0.03891377647717794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,4096,0.012614222036467658
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,4096,0.01828977796766493
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,3584,0.011869333684444427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,4096,0.036421332094404436
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,3584,0.01819022165404426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,3072,0.011001778145631155
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,3584,0.03511022196875678
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,3072,0.016755556066830952
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,2560,0.008433777425024245
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,2560,0.0162782222032547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,3072,0.03436177637841966
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,2048,0.007245333658324347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,2560,0.03352977832158407
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,2048,0.014360888136757744
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,1536,0.006411555740568373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,2048,0.029480000336964924
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,1536,0.014364444547229342
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,1024,0.004266666869322459
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,768,0.003941333128346337
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,1536,0.02903288933965895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,1024,0.013371555341614617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,65536,0.07306577761967976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,1024,0.027243556247817144
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,768,0.013714666995737286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,512,0.0036275556517971885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,768,0.02760355505678389
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,512,0.013481777575280933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,256,0.0032506665835777917
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,512,0.027112000518374976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,256,0.012348444097571902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,128,0.0030088888274298776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,128,0.011520889070298938
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,64,0.003121777748068174
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,256,0.0251226673523585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,32,0.002826666666401757
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,65536,0.1422355572382609
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,128,0.023959111836221483
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,64,0.011705777711338468
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,32,0.011569778124491373
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,16384,0.06164177921083239
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,16384,0.027408889598316614
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,12288,0.023782221807373896
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,16384,0.027898665931489732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,12288,0.05386044581731161
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,12288,0.022416889667510986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,10240,0.019674667053752475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,10240,0.022039999564488728
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,8192,0.016720889343155753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,10240,0.04682222339842054
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,8192,0.02050488856103685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,65536,0.11884444289737278
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,7168,0.015304888288180033
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,65536,0.06757066647211711
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,7168,0.01976711054642995
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,8192,0.046442667643229164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,6144,0.013707555830478668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,6144,0.019204444355434842
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,5120,0.012815111213260226
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,7168,0.04091111156675551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,6144,0.039529777235454984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,5120,0.01871288816134135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,4096,0.011316444310877057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,4096,0.017635555730925668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,5120,0.039889779355790876
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,4096,0.035471111536026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,3584,0.010430222584141625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,3584,0.017303110824690927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,3072,0.009957333405812582
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,2560,0.007791111038790808
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,3584,0.034965333011415266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,3072,0.016196444630622864
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,2560,0.014979556202888489
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,65536,0.15552000204722086
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,2048,0.00685511115524504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,3072,0.033788445923063494
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,2048,0.01425155500570933
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,1536,0.005687110953860813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,2048,0.029362665282355413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,2560,0.03204977843496535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,1536,0.014335110783576965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,1024,0.004311999926964442
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,1024,0.02720888952414195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,1536,0.02903199858135647
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,512,0.003479111111826367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,1024,0.013280889226330651
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,768,0.0038524443904558816
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,768,0.013960889644092984
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,512,0.027277333868874445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,768,0.027422222826215956
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,512,0.013442666994200813
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,256,0.0031946665710873077
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,256,0.012298666768603854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,128,0.0029768887907266617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,256,0.025148444705539282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,65536,0.09648711151546902
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,128,0.011511999699804517
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,64,0.0028008888992998335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,128,0.023966223001480103
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,32,0.0028159999185138275
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,64,0.01147377739350001
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,32,0.011711111499203576
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,16384,0.022517333428064983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,16384,0.026367111338509455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,12288,0.021819555097156104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,16384,0.05814844369888306
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,10240,0.01645155582163069
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,12288,0.051746666431427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,10240,0.020908445119857788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,8192,0.014077333940400017
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,10240,0.04561600089073181
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,8192,0.020374221934212577
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,7168,0.013235555754767524
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,12288,0.018411555223994784
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,7168,0.019516444868511625
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,8192,0.044814222388797335
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,6144,0.012192000117566852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,7168,0.040468444426854454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,6144,0.018778666853904724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,5120,0.01108000013563368
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,6144,0.03903377718395657
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,5120,0.01834844383928511
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,4096,0.009986666341622671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,5120,0.03923111160596212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,4096,0.016862221890025668
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,3584,0.009487110707494948
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,65536,0.06169866853290134
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,4096,0.035113778379228376
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,3584,0.0162453336848153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,3072,0.00905688852071762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,3584,0.034352888663609825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,3072,0.014895111322402954
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,2560,0.007376888559924231
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,3072,0.03204177816708883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,2560,0.014857777290874057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,2048,0.006360888895061281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,65536,0.14234933588239881
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,2048,0.0289066665702396
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,2560,0.030850665436850652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,2048,0.0140666663646698
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,1536,0.004836444639497333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,1024,0.00416355538699362
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,1536,0.014154665999942355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,1536,0.028761777612898085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,1024,0.013215999636385175
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,768,0.003823111040724648
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,1024,0.027049778236283198
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,768,0.013586666848924426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,512,0.0034533333447244433
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,768,0.027502222193611994
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,512,0.013345777988433838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,256,0.0031635556370019913
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,512,0.027095110880004034
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,256,0.024979554944568213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,256,0.012166221936543783
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,128,0.0029164445069101122
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,65536,0.07455999983681573
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,64,0.011397333608733283
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,128,0.011719110939237805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,64,0.002790222151411904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,32,0.00286666676402092
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,128,0.02385155525472429
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,32,0.01145333300034205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,16384,0.023743111226293776
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,12288,0.01423911088042789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,16384,0.05440444416469998
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,12288,0.0214035560687383
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,10240,0.012925333446926542
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,10240,0.020252444677882724
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,12288,0.04816533459557427
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,16384,0.017128000656763714
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,8192,0.01143200033240848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,10240,0.04334222277005514
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,8192,0.019159111711714003
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,7168,0.010955555571450127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,8192,0.042782220575544566
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,7168,0.018431999617152743
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,6144,0.009806222385830348
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,6144,0.017509332961506315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,7168,0.03976622223854065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,65536,0.0503111117415958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,5120,0.009308444129096137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,6144,0.038367112477620445
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,4096,0.008531555533409119
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,5120,0.015903110305468243
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,4096,0.015232889188660515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,5120,0.036045332749684654
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,3584,0.008373333348168267
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,4096,0.03232177760865953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,3584,0.015609777635998197
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,3072,0.0075902218619982404
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,3584,0.031075557072957356
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,3072,0.014690667390823364
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,2560,0.006613333192136552
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,65536,0.12479821840922038
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,3072,0.03031555480427212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,2560,0.01479822231663598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,2048,0.005946666830115848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,2560,0.02956177790959676
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,2048,0.01387288918097814
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,1536,0.004819555415047539
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,65536,0.050365332100126475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,2048,0.028189334604475234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,1024,0.004168888760937585
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,1536,0.014089778065681458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,1536,0.028361777464548748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,768,0.027253333065244887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,1024,0.013152889079517789
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,768,0.0038080000215106537
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,1024,0.026649778087933857
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,512,0.003495111233658261
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,768,0.013625777430004544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,512,0.013312000367376538
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,256,0.0032053333189752367
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,512,0.026704000102149114
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,256,0.012261333564917246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,128,0.002936888900068071
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,256,0.02528088953759935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,128,0.011580444044537015
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,64,0.003101333354910215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,128,0.02383288906680213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,32,0.0030657777355776895
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,64,0.011427555647161273
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,32,0.01146577795346578
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,16384,0.01406044430202908
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,16384,0.022572444544898138
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,12288,0.012356444365448423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,16384,0.05170222123463949
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,12288,0.020594666401545208
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,10240,0.011545778148704104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,12288,0.04604711135228475
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,65536,0.11107733514573838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,10240,0.019816888703240287
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,8192,0.010348444183667501
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,8192,0.01754044493039449
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,10240,0.04159199860360887
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,7168,0.009566222627957663
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,8192,0.041101333167817854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,7168,0.017229333519935608
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,65536,0.04597511225276523
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,6144,0.008743999732865227
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,7168,0.0384168889787462
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,6144,0.016327111257447135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,5120,0.008778666456540426
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,6144,0.03429866830507914
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,5120,0.015857777661747403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,4096,0.007718222008811102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,5120,0.034062223301993474
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,4096,0.015230221880806817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,3584,0.008008889026112026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,4096,0.0307608875963423
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,3584,0.015424888994958667
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,3072,0.007158222297827403
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,3584,0.03033066789309184
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,3072,0.014679110712475248
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,2560,0.014721777704026965
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,2560,0.0063777777055899305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,65536,0.03943822118971083
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,3072,0.029300444655948218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,2048,0.00545866663257281
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,2560,0.028716444969177246
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,2048,0.013971555564138623
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,1536,0.00479644454187817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,2048,0.02776266634464264
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,1536,0.014257777068350049
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,1024,0.004109333372778363
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,1536,0.027816000911924574
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,768,0.003755555384688907
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,1024,0.012998222476906247
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,1024,0.026583110292752583
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,512,0.003504000190231535
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,768,0.013613333304723104
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,512,0.013185777597957186
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,768,0.027112000518374976
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,256,0.003321777822242843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,512,0.026695999834272597
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,256,0.012281777958075205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,128,0.0030666666312350165
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,256,0.024875554773542616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,128,0.011512888802422417
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,64,0.0029608888758553397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,128,0.02371999952528212
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,32,0.0028799999919202593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,64,0.01146488885084788
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,32,0.011493333511882357
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,16384,0.011417778001891242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,16384,0.022270222504933674
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,12288,0.00981244444847107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,16384,0.04387822084956699
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,12288,0.0203448881705602
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,10240,0.009479111267460717
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,10240,0.019440889358520508
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,12288,0.04203111264440748
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,8192,0.008607111043400234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,65536,0.044433779186672635
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,8192,0.016972444123691983
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,7168,0.008626666333940294
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,10240,0.03738844394683838
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,65536,0.10255822208192612
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,8192,0.03636888994110955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,7168,0.016516443755891588
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,6144,0.007631999750932057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,7168,0.033181332879596286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,6144,0.01611022154490153
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,5120,0.007428444094128079
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,6144,0.03169511093033685
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,5120,0.01554666625128852
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,4096,0.00657066661450598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,65536,0.029323554701275293
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,5120,0.032390223609076604
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,4096,0.015218666858143277
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,3584,0.007147555549939473
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,4096,0.02923999892340766
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,3584,0.015112888481881885
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,3072,0.0068400001360310455
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,3584,0.029746668206320867
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,2560,0.01475911173555586
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,3072,0.014712888333532544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,2560,0.006263111200597551
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,3072,0.028259555498758953
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,2048,0.005605333381228977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,2560,0.028500444359249536
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,2048,0.013879111243618859
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,1536,0.004941333499219683
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,2048,0.02739288906256358
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,1024,0.013003555436929068
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,1536,0.014223999447292753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,1024,0.004232888834344016
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,1536,0.02767555581198798
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,768,0.003943110919660992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,1024,0.02641244398223029
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,768,0.013665777941544851
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,512,0.00360000009338061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,512,0.013217777841620974
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,768,0.02703911066055298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,256,0.003286222202910317
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,512,0.02659822172588772
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,256,0.012302222351233164
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,128,0.002999111182159848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,256,0.024860444996092055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,128,0.01145155562294854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,64,0.0028719999310043124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,32,0.0028364445186323593
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,128,0.023804444405767653
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,64,0.01145600030819575
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,32,0.011420444481902652
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,65536,0.017303110824690927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,16384,0.021327111456129286
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,16384,0.008538666698667739
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,65536,0.0440995560752021
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,12288,0.00815288888083564
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,65536,0.09231911102930705
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,12288,0.018976888722843595
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,16384,0.04158666729927063
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,10240,0.007799111306667328
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,12288,0.03712088863054911
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,10240,0.017782222893502977
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,8192,0.006979555719428592
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,10240,0.03518133362134298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,8192,0.016993777619467843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,7168,0.00685777763525645
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,8192,0.0329751107427809
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,7168,0.016501333978441026
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,6144,0.006616000086069107
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,7168,0.03194933467441135
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,5120,0.030082666211658057
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,6144,0.016050666570663452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,5120,0.0069226668112807805
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,6144,0.031360887818866305
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,5120,0.015486222174432544
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,4096,0.006617777877383762
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,4096,0.015056000815497505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,3584,0.006959111326270633
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,4096,0.02923733327123854
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,3072,0.006627555522653792
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,3072,0.028284443749321833
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,3584,0.015173332558737861
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,3584,0.02952444553375244
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,3072,0.014484445254007975
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,2560,0.00611200018061532
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,2560,0.014714666538768344
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,2048,0.005484444399674733
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,1536,0.014031999640994601
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,1024,0.0041644444896115195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,2048,0.013783111340469785
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,2560,0.028547555208206177
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,1536,0.004826666580306159
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,2048,0.027729777826203242
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,768,0.013592888911565145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,1536,0.027600889404614765
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,768,0.003834666477309333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,1024,0.013230221966902414
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,512,0.003452444449067116
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,1024,0.026311111119058397
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,768,0.026889777845806543
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,256,0.003119999956753519
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,512,0.013381332986884646
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,512,0.026564444104830425
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,256,0.012258666257063547
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,128,0.002941333378354708
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,256,0.024811555941899616
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,128,0.011459555890825061
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,64,0.0027786667148272195
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,128,0.023769777682092454
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,32,0.0027928888383838865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,64,0.011552000211344825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,32,0.011416888899273343
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,16384,0.020285333196322124
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,65536,0.011902222202883827
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,65536,0.04319466816054451
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,16384,0.006843555718660355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,65536,0.09285155269834731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,12288,0.006575999988449945
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,16384,0.040778666734695435
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,10240,0.006698666761318843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,12288,0.018631110588709515
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,12288,0.03696799940533108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,10240,0.0177875558535258
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,8192,0.006783111227883234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,10240,0.03506044546763102
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,8192,0.01683466633160909
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,7168,0.006579555571079254
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,8192,0.03301511208216349
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,7168,0.016336888074874878
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,6144,0.006506666541099548
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,7168,0.03222844335767958
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,6144,0.015840000576443143
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,5120,0.00674044465025266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,6144,0.0311119986904992
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,5120,0.015392000476519266
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,4096,0.00657066661450598
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,5120,0.03009422289000617
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,4096,0.015159999330838522
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,3584,0.006751110984219446
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,4096,0.02906222144762675
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,3072,0.014136888914638095
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,3584,0.015005333556069268
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,3072,0.006598222172922558
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,3584,0.029457777738571167
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,2560,0.0058266665372583605
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,3072,0.02838311261600918
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,2560,0.014208000567224292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,2048,0.005218666460778978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,2560,0.02843733297453986
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,2048,0.013719999955760108
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,1536,0.00462311092350218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,2048,0.027223111854659185
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,1536,0.013932444983058505
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,1024,0.003924444317817688
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,1536,0.027495111028353374
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,1024,0.012882666455374824
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,768,0.0037964445849259696
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,1024,0.026650667190551758
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,768,0.013452444639470843
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,512,0.003491555651028951
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,512,0.02650044361750285
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,768,0.02681155502796173
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,512,0.013088888592190213
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,256,0.024728000164031982
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,256,0.003125333330697484
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,256,0.012327999704413943
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,128,0.002951111023624738
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,128,0.011562666959232755
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,64,0.0027662221756246355
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,128,0.02360444433159298
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,32,0.0027706666539112725
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,64,0.011335111326641507
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,32,0.011560888753996955
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,65536,0.009705777797434065
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,16384,0.006461333483457565
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,65536,0.04255377915170458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,12288,0.0064328888224230866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,16384,0.02001333402262794
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,10240,0.006956444846259222
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,12288,0.018255111243989732
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,8192,0.006478222293986215
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,10240,0.0173973325226042
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,7168,0.006396444307433234
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,8192,0.016526222229003906
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,6144,0.0063457778758472866
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,7168,0.016207110550668504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,5120,0.006628444625271692
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,6144,0.015632889337009855
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,4096,0.0063031112982167145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,5120,0.015145777000321282
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,3584,0.006746666712893381
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,4096,0.014762666490342883
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,3072,0.0064560001095136
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,3584,0.014947555131382413
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,2560,0.005780444376998478
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,3072,0.014092443717850579
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,2048,0.005131555514203178
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,2560,0.014127110441525778
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,1536,0.0045422220395671
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,2048,0.013402666482660504
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,1024,0.00397866674595409
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,1536,0.013759999639458127
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,768,0.003620444486538569
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,1024,0.013206221991115145
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,512,0.0032835555159383347
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,768,0.013414222333166333
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,256,0.003087999920050303
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,512,0.013193777865833707
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,128,0.0029297777348094513
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,256,0.012114666402339935
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,32,0.011362666885058085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,64,0.0027928888383838865
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,128,0.01143200033240848
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,64,0.011307555768224927
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,32,0.002790222151411904
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,65536,0.008663111262851292
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,16384,0.006625777731339137
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,65536,0.04158400164710151
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,12288,0.006404444575309753
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,16384,0.020211555891566806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,10240,0.006643555644485686
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,12288,0.01850666602452596
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,8192,0.006518222391605377
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,10240,0.01748533381356133
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,7168,0.00647288892004225
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,8192,0.016727111405796476
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,6144,0.0063279999627007385
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,7168,0.016247999336984422
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,5120,0.006692444284756978
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,6144,0.015747555428081088
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,4096,0.006351999938488007
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,5120,0.015341333217091031
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,3584,0.006771555377377405
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,4096,0.01479111115137736
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,3072,0.0064293332397937775
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,3584,0.0147733340660731
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,2560,0.005793777604897817
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,3072,0.01408266690042284
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,2048,0.005140444470776452
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,2560,0.014089778065681458
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,1536,0.004531555705600315
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,2048,0.013583111266295115
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,1024,0.003932444585694207
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,1536,0.013573333621025085
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,768,0.0036026665733920205
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,1024,0.012921777864297232
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,512,0.003351111171974076
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,768,0.013456888496875763
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,64,0.002711999954448806
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,256,0.003028444531891081
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,512,0.013195555243227216
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,128,0.00282577777074443
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,256,0.012215110990736218
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,128,0.01142133358452055
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,32,0.0027306667632526825
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,64,0.01129333343770769
VLLM,0.12.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,32,0.011531555818186866
