framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,10240,7.856960720486111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,12288,8.999923706054688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,16384,11.650513543023003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,10240,12.903173658582899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,12288,15.340828789605034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,8192,5.939644283718533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,16384,20.659481472439236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,8192,10.206505669487848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,7168,5.284987555609809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,7168,9.00441657172309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,6144,4.6569803025987415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,6144,7.7155507405598955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,5120,6.482041676839192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,5120,5.0684967041015625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,4096,5.155004289415148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,4096,3.7369600931803384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,16384,15.3843502468533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,3584,5.288325415717231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,3584,3.2424276140001087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,3584,3.4473404354519315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,3072,4.36903805202908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,3072,2.7860844930013022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,12288,10.959357367621529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,2560,3.7655296325683594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,10240,9.074570549858942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,2560,2.3134167989095054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,2048,2.65413326687283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,2048,1.8477066887749567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,3072,2.7559466891818576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,1536,2.3477564917670355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,1536,1.3869155248006184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,2560,2.520849863688151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,1024,1.689652442932129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,1024,0.9335412979125977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,8192,9.953052096896702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,768,1.790524376763238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,2048,1.8644720713297527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,768,0.7332800229390463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,7168,6.433428446451823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,512,0.5701644685533311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,512,1.3409386740790472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,256,0.8366186883714465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,256,0.45909865697224933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,6144,5.603292253282334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,128,0.4417262342241075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,128,0.4366933239830865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,5120,4.781623840332031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,64,0.4112079938252767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,64,0.4149048858218723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,65536,32,0.4002266724904378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,65536,32,0.40730932023790145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,1536,1.4831760194566515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,4096,3.7754105461968313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,256,0.5121742354498969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,512,0.6351564195421007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,768,0.849462244245741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,1024,1.0560195710923936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,16384,3.922339545355903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,12288,3.0158763461642795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,16384,9.476914299858942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,12288,4.090762668185763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,10240,2.541376961602105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,65536,21.01719495985243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,65536,18.651024712456596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,10240,4.585365295410156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,8192,1.994265874226888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,8192,2.731756422254774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,7168,1.785027609931098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,7168,2.349735895792643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,6144,2.2477128770616317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,6144,1.5825031068589952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,16384,4.8520630730523004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,5120,1.8436133066813152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,12288,3.7694846259223094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,5120,1.50273863474528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,4096,1.3219973246256511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,10240,3.0940089755588107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,4096,1.1944657431708443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,8192,2.873226589626736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,3584,1.1651360193888347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,3584,1.0521787007649739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,65536,20.249043782552082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,3072,1.2961822085910373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,3072,0.9076773325602213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,7168,2.15339109632704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,2560,0.8180738025241427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,2560,0.7598942120869955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,2048,0.690034654405382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,2048,0.608858691321479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,6144,1.842715581258138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,1536,0.5465102195739746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,1536,0.4586168924967448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,65536,128,0.4727217886182997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,3584,1.0774755477905273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,1024,0.37017422252231175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,5120,1.5397698084513347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,1024,0.30976088841756183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,768,0.30316087934705943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,768,0.23160799344380698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,512,0.20911110772026908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,4096,1.2388053470187717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,2560,0.8196018007066516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,512,0.17620533042483857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,256,0.12759021917978922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,256,0.1364426612854004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,128,0.11513333850436741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,128,0.12227733929951985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,3072,0.9606648551093208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,64,0.107040888733334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,1536,0.5091830889383951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,64,0.11569155587090386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,16384,32,0.1041075587272644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,16384,32,0.11572799417707656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,2048,0.6498648855421278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,65536,15.222471449110243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,256,0.16195289293924967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,65536,18.452000935872395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,512,0.22822043630811903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,16384,4.183908250596788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,16384,3.3464452955457897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,12288,3.1192764706081815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,768,0.2991982301076253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,12288,2.562948438856337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,10240,2.446549309624566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,10240,2.1342425876193576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,8192,1.946569866604275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,8192,1.69196351369222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,1024,0.3711555533938938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,7168,2.051216973198785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,7168,1.5055902269151475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,6144,1.5017324023776586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,6144,1.3045057720608182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,65536,17.08969963921441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,5120,1.490812407599555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,5120,1.2464417351616752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,16384,6.407755533854167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,12288,3.108055962456597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,10240,2.5610018836127386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,4096,0.990937762790256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,4096,2.743529849582248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,3584,0.8925946553548177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,8192,2.0496383243136935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,3072,0.7543004353841146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,3584,2.846028433905708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,3072,0.753053347269694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,7168,1.79949951171875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,2560,0.628161801232232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,2560,0.6331955591837565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,2048,0.5081653594970703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,2048,0.5065955585903591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,6144,1.5399804645114477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,1536,0.400926219092475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,1536,0.38167378637525773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,16384,128,0.1439137723710802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,1024,0.2646755642361111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,1024,0.25877155197991264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,3584,0.9238559934828016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,768,0.21878577603234184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,768,0.1932764450709025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,5120,1.2866729100545247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,512,0.15994400448269314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,512,0.14059200551774767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,4096,1.0381146536933052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,256,0.1009662217564053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,256,0.1252773337894016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,2560,0.6727671093410916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,128,0.08709244595633613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,128,0.09697688950432672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,3072,0.8253662321302625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,64,0.07992888821495904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,64,0.1013537777794732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,12288,32,0.07882489098442926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,12288,32,0.09300888909233941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,1536,0.4359742270575629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,2048,0.5531048774719238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,65536,12.564556545681425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,256,0.1349031130472819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,512,0.19155200322469076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,16384,3.4700096978081594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,16384,3.0092942979600696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,65536,19.897872924804688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,768,0.2505768934885661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,12288,2.644218656751845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,12288,2.3176337348090277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,1024,0.3108168972863091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,10240,2.2321599324544272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,10240,1.909488042195638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,8192,1.686465793185764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,8192,1.5341555277506511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,7168,1.465026643541124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,7168,1.3688355551825628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,6144,1.3799538082546656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,6144,1.1954978307088215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,65536,15.416022406684029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,5120,1.1470186445448134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,16384,3.7383367750379772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,5120,1.1087652842203777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,4096,0.928959952460395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,12288,3.7076623704698353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,10240,2.3211004469129777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,4096,0.8865262137518989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,8192,1.8690853118896484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,3584,0.7194515334235297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,3584,0.7783519956800672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,3072,0.6727324591742622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,3072,0.6709573533799914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,7168,1.618559095594618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,2560,0.544906669192844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,2560,0.5605306625366211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,2048,0.4434631135728624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,2048,0.45108620325724286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,6144,1.4031635920206706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,1536,0.34410844908820254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,1536,0.34339290195041233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,12288,128,0.12036444081200494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,1024,0.2362684408823649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,3584,0.8381173345777723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,5120,1.1708240509033203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,1024,0.23239644368489584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,768,0.18938310941060385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,768,0.1736853387620714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,512,0.1378693315717909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,512,0.1254026624891493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,4096,0.9575387107001411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,2560,0.6140444543626573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,256,0.0958026647567749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,256,0.09710133075714111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,128,0.07493155532413058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,128,0.0837084452311198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,1536,0.3946586714850531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,64,0.06915999783409967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,64,0.08078755272759332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,10240,32,0.06858222352133857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,3072,0.7255502276950412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,10240,32,0.08046666781107585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,2048,0.4996044370863173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,65536,10.855705261230469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,65536,11.759925842285156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,16384,2.8924800025092234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,16384,2.6857723659939237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,12288,2.0382088555230036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,256,0.12128089533911811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,512,0.1711946725845337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,10240,1.7733084360758464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,12288,2.041647169325087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,10240,1.7158880233764648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,768,0.22939022382100424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,8192,1.37371826171875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,8192,1.3646649254692926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,7168,1.2427049212985568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,7168,1.234389305114746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,1024,0.2833706537882487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,6144,1.1379938125610352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,6144,1.0755884382459853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,65536,13.391827053493925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,16384,3.342224968804253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,5120,0.8678044213189019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,5120,0.9599857330322266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,4096,0.6563084390428331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,4096,0.7666444248623319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,12288,2.9591973622639975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,10240,2.0856711069742837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,3584,0.666536013285319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,3584,1.1342764960394967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,3072,0.9393146302964953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,3072,0.5665013525221083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,8192,1.6596284442477758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,2560,0.45256710052490234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,2560,0.4722897741529677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,7168,1.4537137349446614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,2048,0.36643822987874347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,2048,0.37149779001871747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,6144,1.2604684829711914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,1536,0.27958665953742134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,1536,0.2788417869144016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,3584,0.7495475345187717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,1024,0.19278489218817818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,1024,0.18822399775187174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,5120,1.0442302491929796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,768,0.14587555991278753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,768,0.14271467261844212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,4096,0.8546204566955566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,512,0.11015822490056355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,512,0.10779466893937852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,2560,0.5529333220587836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,256,0.06774222188525729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,256,0.07800622118843926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,10240,128,0.10582577519946629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,128,0.05957333246866862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,128,0.06758044163386027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,3072,0.6556551191541884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,64,0.05478044350941976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,64,0.06384622388415866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,8192,32,0.05388622151480781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,8192,32,0.06579199764463636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,1536,0.35628355873955625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,2048,0.4525369008382161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,65536,12.600322299533419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,65536,10.28739505343967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,16384,2.5080880059136286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,256,0.10466488864686753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,16384,5.915686289469401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,128,0.0910044444931878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,512,0.15182488494449192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,12288,1.9480088551839192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,10240,1.4671484629313152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,768,0.2006995545493232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,10240,1.6154372957017686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,12288,4.196691724989149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,8192,1.0440195931328666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,8192,1.2819218105740018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,7168,0.9895413716634115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,7168,1.136675516764323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,6144,0.7783573468526205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,6144,0.986055162217882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,8192,1024,0.2538995477888319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,5120,0.7473102145724826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,5120,0.9029191335042318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,65536,12.663201226128471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,4096,0.5995555453830296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,4096,0.7168604532877604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,16384,3.124894248114692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,3584,0.5478782123989528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,12288,2.3530036078559027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,3584,0.6265075471666124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,3072,0.44557423061794704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,10240,1.9472577836778429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,3072,0.533713764614529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,2560,0.3608080016242133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,8192,2.1144462161593967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,2560,0.4452551205952962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,7168,1.3668301900227864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,2048,0.2843671109941271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,2048,0.3528409004211426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,1536,0.2170871098836263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,1536,0.26327821943495006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,6144,1.1885813607109916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,1024,0.1616764465967814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,1024,0.1782311068640815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,5120,0.9939875072903104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,768,0.13032533062828913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,768,0.13827289475335017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,3584,0.7170071072048612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,512,0.09721244706047906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,512,0.10214400291442871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,4096,0.8095448811848959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,256,0.061234666241539844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,256,0.07405777772267659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,2560,0.5280497868855795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,128,0.05386311146948072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,128,0.061850666999816895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,1536,0.3360133435991075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,64,0.05275377631187439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,64,0.059373332394493945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,7168,32,0.04831111099984911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,7168,32,0.06020266479916043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,3072,0.6189457575480143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,2048,0.43172356817457413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,768,0.1892933315700955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,16384,2.0016443464491105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,16384,2.3530356089274087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,65536,7.763274298773871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,1024,0.24072533183627656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,12288,1.7261892954508464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,12288,1.7880880567762587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,65536,9.777557373046875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,10240,1.2662159601847331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,10240,1.4877600140041773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,8192,0.9814258151584201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,8192,1.2100195354885526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,256,0.09792977571487427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,512,0.1430506706237793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,7168,0.8851280212402344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,7168,1.0654470655653212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,6144,0.708615091111925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,6144,0.9231999715169271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,16384,2.915492375691732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,5120,0.6836382018195258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,12288,2.181272930569119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,5120,0.836948447757297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,10240,1.8232754601372614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,4096,0.5481635729471842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,4096,0.6663351058959961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,3584,0.4378382099999322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,65536,11.796463012695312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,3584,0.5796675682067871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,3072,0.3711804548899333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,3072,0.497869332631429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,8192,1.4791733423868816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,2560,0.34031467967563206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,2560,0.41433244281344944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,7168,1.2921786838107638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,2048,0.26619201236301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,2048,0.32761867841084796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,6144,1.1249706480238173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,1536,0.20561067263285318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,1536,0.24516712294684517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,3584,0.6723448965284559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,1024,0.14600622653961182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,1024,0.16646577252282038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,5120,0.929441769917806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,4096,0.7584337658352323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,768,0.11326932907104492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,768,0.12701333893669978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,2560,0.49656089146931964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,512,0.08391022019916111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,512,0.09470222393671672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,256,0.0530008872350057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,256,0.06796622276306152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,3072,0.583114677005344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,128,0.04794044295946757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,128,0.05630489190419515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,1536,0.31894667943318683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,64,0.04373777906099955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,64,0.05369866556591458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,6144,32,0.04219288958443535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,6144,32,0.05417689018779331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,2048,0.40635021527608234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,7168,128,0.08425777488284641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,65536,6.684080759684245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,16384,1.869217766655816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,65536,8.894995795355902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,16384,2.203991148206923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,12288,1.318746672736274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,12288,1.6657270855373805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,256,0.09179555707507664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,128,0.07795377572377522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,10240,1.0812835693359375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,10240,1.4002008438110352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,512,0.1323608954747518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,8192,0.8162124421861437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,8192,1.1239066653781467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,7168,0.7498115433586968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,768,0.1784657769733005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,7168,1.0098649130927193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,6144,0.6933226585388184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,6144,0.8714977900187174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,6144,1024,0.22467645009358725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,5120,0.5370702213711208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,5120,0.7684239811367459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,16384,2.7248452504475913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,4096,0.41645333502027726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,4096,0.6181048817104763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,12288,2.0801654391818576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,65536,18.64141845703125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,3584,0.38141510221693253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,3584,0.5870169003804525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,3072,0.4347208870781793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,3072,0.5226524670918783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,10240,1.7139164606730144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,2560,0.27431199285719127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,2560,0.3874008920457628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,8192,1.3889040417141383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,2048,0.22901066144307455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,2048,0.3050897651248508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,7168,1.2152026494344075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,1536,0.1712426609463162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,1536,0.22766221894158256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,6144,1.0536550945705838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,1024,0.12056177192264134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,1024,0.15624622503916422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,5120,0.8769911130269369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,768,0.09557333257463242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,768,0.11762221654256184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,3584,0.633770677778456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,512,0.07204977671305339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,4096,0.712681770324707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,512,0.08634310960769653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,256,0.04870133267508613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,256,0.06240089072121514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,2560,0.4649804433186849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,128,0.039876444472206965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,128,0.05007822314898173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,1536,0.2986808882819282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,64,0.03608355588383145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,64,0.04828800095452202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,5120,32,0.035820444424947105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,5120,32,0.047712001535627574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,3072,0.5508640077379015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,2048,0.382405334048801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,65536,5.190148247612847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,768,0.16432532999250624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,16384,1.4205492867363825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,16384,2.0344115363226996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,1024,0.2115537722905477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,12288,0.9762907028198242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,65536,9.768115573459202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,12288,1.541042645772298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,10240,1.2947048611111112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,10240,2.9897083706325955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,8192,2.9594870673285594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,256,0.0845786664221022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,128,0.07131555345323351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,7168,0.5660710864596897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,8192,1.9430825975206163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,5120,512,0.12276266680823432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,7168,0.910594622294108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,6144,0.4870719909667969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,6144,0.7845502429538302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,5120,0.3767057789696588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,5120,0.7008488972981771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,16384,2.5455485449896917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,4096,0.33571378389994305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,4096,0.5651955604553223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,12288,1.9090044233534071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,3584,0.2922426594628228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,3584,0.49199289745754665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,10240,1.602110226949056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,65536,10.23395029703776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,3072,0.25357866287231445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,3072,0.42226754294501406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,2560,0.21397866143120658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,2560,0.35628533363342285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,8192,1.302477306789822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,2048,0.17918399969736734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,2048,0.28116265932718915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,7168,1.2562221950954862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,6144,0.9770381715562609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,1536,0.13878844843970403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,1536,0.21170400248633492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,1024,0.09621333413653904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,1024,0.14237511157989502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,5120,0.822380436791314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,768,0.07525600327385797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,768,0.10637955533133613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,3584,0.5889066590203179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,512,0.0566542214817471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,512,0.07821155256695218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,2560,0.4346764352586534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,256,0.037439998653199934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,4096,0.6694489055209689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,256,0.054261333412594266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,128,0.03276088833808899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,128,0.04353688822852241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,3072,0.5121546851264106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,64,0.029602666695912678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,64,0.04159377680884467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,4096,32,0.0299253331290351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,4096,32,0.04108177622159322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,1536,0.27782400449117023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,2048,0.35853510432773167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,65536,4.576683468288846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,768,0.15305156177944607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,16384,1.261807123819987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,16384,1.9424336751302083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,1024,0.19649777147505018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,12288,0.7763555314805773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,12288,1.478294266594781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,65536,9.976566738552517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,10240,0.6974239879184299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,10240,1.5612604353162978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,8192,1.6221324072943792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,8192,1.3032577302720811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,128,0.06371021933025785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,256,0.07711644305123223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,4096,512,0.11223822169833714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,7168,0.5366524590386285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,6144,0.4643102222018772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,7168,0.8712924321492513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,6144,0.7501964569091797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,5120,0.3737528853946262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,5120,0.6756773524814181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,16384,2.455895105997721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,4096,0.30079200532701283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,4096,0.5399484634399414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,12288,1.857927958170573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,3584,0.2595377763112386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,3584,0.47383642196655273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,10240,1.5522346496582031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,3072,0.22103377183278403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,3072,0.40654577149285215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,65536,9.983770582411024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,2560,0.18999200397067598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,2560,0.34251822365654844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,8192,1.595070203145345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,2048,0.4219591087765164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,2048,0.2814835442437066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,7168,1.0919378068712022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,1536,0.11665866110059951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,1536,0.20538132720523408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,6144,0.9462853537665473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,1024,0.08405955632527669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,1024,0.1401431163152059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,5120,0.7951279746161567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,768,0.06667555703057183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,768,0.10325866937637329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,3584,0.5714337560865614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,512,0.05037600133154127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,512,0.07507199711269803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,4096,0.6461608674791124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,2560,0.42334755261739093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,256,0.03401866555213928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,256,0.052617778380711876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,128,0.02947288751602173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,128,0.04128977656364441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,3072,0.49528710047403973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,64,0.02741777731312646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,64,0.03864266806178623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3584,32,0.02699288891421424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3584,32,0.038262221548292376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,1536,0.2671884430779351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,768,0.14575200610690647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,2048,0.3466773298051622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,65536,4.033003489176433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,16384,1.9534195793999567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,16384,1.8693663279215496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,1024,0.18872443834940592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,12288,0.7240791320800781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,12288,1.423349380493164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,65536,7.494825575086806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,10240,0.6301120122273763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,10240,1.1885751088460286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,8192,0.5154817899068197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,8192,0.9658311208089193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,512,0.10694844192928737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,128,0.059382220109303795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,7168,0.4467484686109755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3584,256,0.07352355453703138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,7168,0.8374266624450684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,6144,0.37609243392944336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,6144,0.7214106453789605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,5120,0.3089946640862359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,5120,0.6412151124742296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,16384,3.5187839931911893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,4096,0.7302622265285916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,4096,0.5286293559604222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,12288,1.782107459174262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,3584,0.21889244185553658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,3584,0.45258580313788516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,10240,1.4933030870225694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,3072,0.19448533323076037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,65536,9.486319647894964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,3072,0.3896213372548421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,2560,0.16004889541202122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,2560,0.328438229031033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,8192,1.203113767835829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,2048,0.1349440018335978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,2048,0.2622426615820991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,7168,1.0593573252360027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,1536,0.10413777828216553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,6144,0.9133947160508898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,1536,0.19879644446902803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,1024,0.07350311014387342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,1024,0.13515733347998724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,5120,0.7703573438856336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,768,0.05891022417280409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,768,0.09868088695738052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,3584,0.5506951014200846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,512,0.04451644420623779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,512,0.07088177733951144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,4096,0.6259724299112955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,256,0.030444443225860596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,2560,0.4066630999247233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,256,0.048623998959859215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,128,0.025868443979157343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,128,0.03810666667090522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,3072,0.4779689047071669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,64,0.0236835562520557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,64,0.03574133250448439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,3072,32,0.024923556380801733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,3072,32,0.03569866551293267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,1536,0.25810133086310494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,2048,0.33237600326538086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,768,0.14127200179629856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,16384,0.8570453325907389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,65536,3.5205627017550998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,16384,1.792451646592882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,12288,0.5987671216328939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,12288,1.3577102025349934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,65536,7.160335116916233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,1024,0.1824044386545817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,10240,0.5373582310146755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,10240,1.134940465291341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,8192,0.42136798964606387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,8192,0.9087759653727213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,128,0.05628710985183716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,7168,0.6943128903706869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,256,0.0703386664390564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,3072,512,0.10170400142669678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,6144,0.3005582226647271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,7168,0.804107560051812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,6144,0.6874391237894694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,5120,0.2584071159362793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,5120,0.6104435390896267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,16384,2.2866569095187717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,4096,0.20845422479841444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,4096,0.49063910378350156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,10240,1.4428577423095703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,12288,1.7182097964816625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,3584,0.18265867233276367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,65536,9.131721496582031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,3584,0.4295670986175537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,3072,0.1593973371717665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,8192,1.164671156141493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,3072,0.3707955678304036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,2560,0.13822933038075766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,2560,0.31248532401190865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,2048,0.11341333389282227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,2048,0.24998132387797037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,7168,1.0194675657484267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,6144,0.8850008646647135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,1536,0.0877679983774821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,5120,0.7403750949435763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,1536,0.191466662618849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,1024,0.06282044119305082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,1024,0.1299120055304633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,768,0.05084355672200521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,768,0.09485599729749893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,3584,0.5316444502936469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,4096,0.6014977561102973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,512,0.03911911116706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,2560,0.3881315655178494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,512,0.06669955783420138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,256,0.027757333384619817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,256,0.04564266734653049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,3072,0.4598577817281087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,128,0.021920000513394673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,128,0.03535111082924737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,64,0.019687111179033916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,1536,0.24686045116848418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,2048,0.32003643777635365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,64,0.0326551099618276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2560,32,0.020337777005301583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2560,32,0.03233777814441257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,768,0.1339555581410726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,65536,2.5807732476128473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,1024,0.17395644717746309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,16384,1.122776879204644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,16384,1.7176435258653429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,12288,1.8617778354220922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,65536,6.870349460177952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,12288,1.3025955624050565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,10240,0.4030124346415202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,10240,1.0850764380560982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,8192,0.340390231874254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,256,0.06623111168543498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,512,0.09678755866156684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,8192,0.8714906904432508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,7168,0.2836222118801541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,7168,0.7674702008565267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,6144,0.2468470997280545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,6144,0.661736011505127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,16384,2.1874311235215926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,5120,0.20554310745663115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,12288,1.9662585788302953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,5120,0.5782471232944065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,65536,8.775905185275608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,4096,0.1716008848614163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,4096,0.46593512429131406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,10240,1.3815173043145075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,3584,0.15176355838775635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,3584,0.4093288845486111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,8192,1.115004433525933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,3072,0.1272453334596422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,3072,0.35314488410949707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,2560,0.10746577713224624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,7168,0.9801208708021375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,2560,0.2972799936930339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,2048,0.08948622147242229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,2048,0.23953689469231498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,6144,0.847591135236952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,1536,0.06956533590952556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,1536,0.18252000543806288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,3584,0.5111982027689616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,5120,0.7127360237969292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,1024,0.05062222149637011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2560,128,0.05326133304172092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,1024,0.12355200449625652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,768,0.041365332073635526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,768,0.08936888641781277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,4096,0.576416015625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,2560,0.3746684392293294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,512,0.031467556953430176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,3072,0.4389342202080621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,512,0.06292711363898383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,256,0.02179199953873952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,256,0.04215111003981697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,1536,0.23751645618014863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,128,0.017024000485738117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,128,0.03217688865131802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,2048,0.3052266703711616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,64,0.014852444330851236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,64,0.029489778810077246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,2048,32,0.01572800013754103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,2048,32,0.028492444091373022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,65536,2.026096979777018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,16384,0.48706579208374023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,16384,1.651759147644043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,65536,6.5902150472005205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,12288,0.3718239996168349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,12288,1.2530888451470268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,512,0.09201866388320923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,256,0.06211821900473701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,768,0.1269964509540134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,10240,0.30801065762837726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,128,0.049514667855368726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,10240,1.0440915425618489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,8192,0.25670933723449707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,8192,0.8360515700446235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,7168,0.21471644772423637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,2048,1024,0.16733511288960776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,7168,0.7360106574164497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,6144,0.17301866743299696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,6144,0.6330693562825521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,5120,0.16061600049336752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,5120,0.5469164318508571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,16384,2.1014141506618924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,4096,0.12566667132907444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,4096,0.4417137834760878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,65536,12.850971645779081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,12288,1.5835786395602758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,3584,0.2264826562669542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,3584,0.3878488805558946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,10240,1.3281653722127278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,3072,0.09932266341315375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,3072,0.33528712060716415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,2560,0.0829226639535692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,8192,1.0707217322455513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,2560,0.28302134407891166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,2048,0.06965333223342896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,2048,0.22774222162034777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,7168,0.9479529062906901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,1536,0.05407289001676771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,6144,0.8122000164455838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,1536,0.1738799942864312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,1024,0.04007110993067423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,1024,0.11739110946655273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,5120,0.6846008830600314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,768,0.032955557107925415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,768,0.08419999811384414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,3584,0.49003288480970597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,512,0.025203555822372437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,512,0.0599271125263638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,4096,0.55312532848782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,256,0.017822222577200997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,256,0.039509332842297025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,2560,0.35866578420003253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,128,0.013745778136783175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,128,0.02934933371014065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,3072,0.4240550994873047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,64,0.011766222616036734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,64,0.025079111258188885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1536,32,0.012358222570684222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1536,32,0.02439911166826884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,65536,1.3344390657212999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,1536,0.22810755835639107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,2048,0.2958915498521593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,16384,0.32661332024468315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,16384,1.578472031487359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,768,0.12071377701229519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,12288,0.25034578641255695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,65536,6.242748260498047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,12288,1.1861822340223525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,10240,0.21210132704840767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,1024,0.1585573355356852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,10240,0.992185804578993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,8192,0.159642669889662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,8192,0.7962746620178223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,256,0.05964533487955729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,7168,0.13992267184787327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,512,0.08860711256663005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,7168,0.7008524470859103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,6144,0.12756800651550293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,6144,0.6043066448635525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,16384,2.0187405480278864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,5120,0.1034266683790419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,5120,0.5152577824062771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,12288,1.5275280210706923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,65536,7.9653074476454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,4096,0.08764621946546768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,4096,0.4161706765492757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,3584,0.07493778069814046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,3584,0.366500457127889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,10240,1.2747155295477974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,3072,0.06624266836378309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,3072,0.3164088990953234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,8192,1.030658615960015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,2560,0.05438399977154202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,2560,0.26665867699517143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,7168,0.9043991300794813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,2048,0.0483511123392317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,2048,0.21552533573574492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,6144,0.7764364348517524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,1536,0.03543466660711501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,1536,0.16455645031399196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,3584,0.46911202536688906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,1024,0.025693333811230127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,1024,0.11093333032396104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,5120,0.6539120144314235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,4096,0.5309484269883897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,768,0.020959999826219346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,768,0.0792435540093316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,512,0.01575555569595761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,2560,0.3408844470977783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,512,0.055128888951407544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,256,0.011190222369299995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,256,0.03254222207599216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,3072,0.40672887696160215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,128,0.008692444198661381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,128,0.023856000767813787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1536,128,0.046259555551740855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,64,0.00756177802880605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,64,0.020958221620983545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,1024,32,0.008099555969238281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,1024,32,0.019861333900027804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,1536,0.21675911214616564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,65536,1.0322959687974718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,2048,0.2836257881588406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,16384,0.26870844099256724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,16384,1.5734418233235676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,65536,6.2010985480414496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,12288,0.18921955426534018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,12288,1.1800959904988606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,256,0.051487111383014254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,10240,0.15887644555833605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,10240,0.9886471430460612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,512,0.08263022369808622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,128,0.039356443617079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,8192,0.12661688857608372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,8192,0.7944489055209689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,7168,0.11712800131903754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,7168,0.6986408763461642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,768,0.11465156078338623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,6144,0.09488355451160008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,6144,0.6020035743713379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,1024,1024,0.15104356076982287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,5120,0.08046399884753756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,5120,0.500225755903456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,16384,2.032663133409288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,4096,0.06873599688212077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,4096,0.40387733777364093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,65536,7.861727396647136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,3584,0.059993777010175914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,3584,0.35555820994906956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,12288,1.4919573465983074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,3072,0.05000888970163134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,3072,0.3062702284918891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,10240,1.2475919723510742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,2560,0.04418311185306973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,2560,0.2589662339952257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,8192,1.0011617872450087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,2048,0.0351528889603085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,2048,0.20970666408538818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,7168,0.879466692606608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,1536,0.027336888843112524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,1536,0.16012267271677652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,6144,0.7601893213060168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,1024,0.020021332634819877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,1024,0.10987555318408543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,5120,0.6377422014872233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,768,0.016192000773217943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,768,0.07886844211154513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,3584,0.45975467893812394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,512,0.012664889295895895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,512,0.05294933252864414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,4096,0.5200071334838867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,256,0.008947555389669206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,256,0.03094666533999973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,3072,0.39760533968607586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,128,0.007229333122571309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,128,0.02243822150760227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,2560,0.3344568941328261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,64,0.006341333190600078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,64,0.019686222076416016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,768,32,0.00648266656531228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,768,32,0.018486221631368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,65536,0.654329776763916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,1536,0.2108880016538832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,2048,0.27480800946553546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,16384,0.17229510678185356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,768,0.11050489213731553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,16384,1.4853537877400715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,65536,5.896209716796875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,12288,0.1290186643600464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,12288,1.1225324206882052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,10240,0.10630222161610921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,10240,0.9401608573065864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,1024,0.1461520062552558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,8192,0.08364355564117432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,8192,0.7532577514648438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,256,0.05063644382688734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,7168,0.08154666423797607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,128,0.037045333120558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,7168,0.6626835399203831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,6144,0.06619022289911906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,6144,0.5707466867234972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,768,512,0.0811048878563775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,5120,0.054930667082468666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,5120,0.4837795363532172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,16384,1.9346879323323567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,12288,1.4632142384847004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,4096,0.04547377758555942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,4096,0.39078132311503094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,3584,0.04024444354905023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,10240,1.2236542171902127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,3584,0.34399644533793133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,3072,0.035559111171298556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,65536,7.632450527615017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,8192,0.9816462198893229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,3072,0.29744089974297416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,2560,0.030361778206295435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,2560,0.25059376822577584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,2048,0.025120000044504803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,2048,0.20246399773491752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,7168,0.8668053415086535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,1536,0.019695111446910434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,1536,0.15512888961368138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,6144,0.7475866741604276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,1024,0.01388088862101237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,1024,0.10616711113188003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,5120,0.6271200180053711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,768,0.011889778077602386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,768,0.07534666856129964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,3584,0.4487324290805393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,512,0.009488888912730748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,512,0.04998133248753018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,2560,0.326566219329834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,256,0.007305777735180325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,4096,0.5072462293836806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,256,0.029197331931855943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,128,0.006080889039569431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,128,0.020837333467271592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,3072,0.3896195623609755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,64,0.005471999860472149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,64,0.018087110585636564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,512,32,0.005623999983072281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,512,32,0.01699199941423204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,2048,0.26760978168911403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,65536,0.4089759985605876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,1536,0.20615466435750326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,16384,0.10154933399624294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,16384,1.4803715811835394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,768,0.10876533057954575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,65536,5.847598181830512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,12288,0.07838577694363065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,12288,1.1134559843275282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,10240,0.06631466415193346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,1024,0.1424071126514011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,10240,0.9316790898640951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,8192,0.05370755659209358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,8192,0.7485769059922961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,128,0.03606044583850437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,256,0.04943289028273689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,7168,0.04761244522200691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,512,512,0.0791813333829244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,7168,0.658516459994846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,6144,0.04222844375504387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,6144,0.5672346750895182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,5120,0.03599822190072801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,5120,0.46952268812391496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,16384,1.9111910925971136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,4096,0.02980088856485155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,65536,7.567711724175347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,4096,0.37963112195332843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,12288,1.439336034986708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,3584,0.026966222458415564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,3584,0.33394845326741535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,3072,0.02401688860522376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,10240,1.205842653910319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,3072,0.28828710979885525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,2560,0.020632889535692003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,2560,0.24299910333421496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,8192,0.9653156068589953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,2048,0.013977777626779346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,2048,0.19680355654822457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,7168,0.8492346869574653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,6144,0.7327946556939019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,1536,0.010797333386209277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,1536,0.1501902209387885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,1024,0.008386666576067606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,1024,0.10225866900549994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,5120,0.6153955459594727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,768,0.0073315559162033936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,768,0.07242488861083984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,4096,0.49974754121568465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,3584,0.4398755497402615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,512,0.006246222390068903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,512,0.048327111535602145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,256,0.005063111169470681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,3072,0.3819315433502197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,256,0.028041776683595445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,128,0.004448889030350579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,2560,0.32050222820705837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,128,0.019527110788557265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,64,0.00406666679514779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,2048,0.26122845543755424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,64,0.016723554995324876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,256,32,0.004287999951177173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,1536,0.20012177361382377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,256,32,0.015607110328144498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,1024,0.13888800144195557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,65536,0.39703822135925293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,768,0.10446666346655951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,16384,0.10360799895392524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,16384,1.4792035420735676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,12288,0.07111555337905884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,65536,5.849347432454427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,12288,1.1147804260253906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,10240,0.06978578037685819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,10240,0.9326764212714301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,512,0.0770346654786004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,256,0.046929776668548584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,8192,0.04836710956361559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,7168,0.04300533400641548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,8192,0.7494604322645398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,256,128,0.03455288873778449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,7168,0.6585591104295518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,6144,0.038541333542929754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,6144,0.5677270889282227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,5120,0.032646222247017756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,5120,0.46834754943847656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,16384,1.8900932735866969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,4096,0.02996888756752014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,12288,1.4255342483520508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,4096,0.3776800102657742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,3584,0.02388977838887109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,3584,0.3328631189134386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,8192,0.9632346895005969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,10240,1.1928461922539604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,3072,0.021341333786646526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,3072,0.287970675362481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,7168,0.8458799786037869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,2560,0.016819554898473952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,2560,0.24281154738532174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,65536,7.48055436876085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,2048,0.009019555317031013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,6144,0.727400885687934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,2048,0.2074835565355089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,1536,0.007530666887760162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,1536,0.1499911149342855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,1024,0.005946666830115848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,1024,0.10176000330183242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,5120,0.612226645151774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,768,0.005355555564165115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,768,0.0726791090435452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,4096,0.4935235447353787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,512,0.004648888690604104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,512,0.04801511102252536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,3584,0.43479466438293457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,256,0.004080888711743885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,256,0.02792888879776001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,3072,0.377199994193183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,128,0.003573333223660787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,128,0.01928622192806668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,2560,0.3174693319532606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,2048,0.25879910257127553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,64,0.0033537778589460584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,64,0.016134222348531086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,128,32,0.003650666524966558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,128,32,0.014845333165592618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,1536,0.1985751125547621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,65536,0.36460622151692706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,16384,0.09017599953545465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,12288,0.0692159997092353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,1024,0.13712178336249456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,16384,1.4781004587809246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,10240,0.05812178055445353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,12288,1.1141315036349826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,8192,0.04737066560321384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,10240,0.9321973588731555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,7168,0.04203111264440748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,8192,0.7487777603997124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,6144,0.03747377792994181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,7168,0.6581857999165853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,65536,5.855082617865668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,6144,0.566706657409668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,5120,0.0314302212662167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,5120,0.46918312708536786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,768,0.10372710890240139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,4096,0.029395557112163965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,3584,0.3340657816992865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,4096,0.379639122221205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,3584,0.023229334089491103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,3072,0.0206631107462777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,3072,0.2889253298441569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,2048,0.1975333293279012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,2560,0.0151573336786694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,2560,0.24345511860317656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,2048,0.009370666411187913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,1024,0.006164444403515921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,256,0.04689866635534498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,512,0.07574933105044894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,768,0.07265244589911567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,1536,0.007840889195601145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,1536,0.15074043803744847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,1024,0.10177866617838542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,768,0.005364444520738389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,512,0.004504000147183736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,512,0.04798577891455757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,256,0.0037342223028341928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,256,0.02730400032467312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,128,0.003370666669474708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,128,0.019295111298561096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,64,0.0032311110860771606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,64,0.016023111012246873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,64,32,0.0032720000793536506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,64,32,0.014873777826627096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,65536,0.3588106632232666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,16384,1.4796115027533636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,16384,0.08910222185982598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,12288,0.06837689214282565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,10240,0.05751644240485298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,12288,1.1139858033921983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,8192,0.7488266626993815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,8192,0.04671288861168755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,10240,0.9320453008015951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,65536,5.855929056803386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,5120,0.030834668212466772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,7168,0.04136888848410712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,4096,0.37927643458048504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,7168,0.6580853462219238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,6144,0.0365866654449039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,6144,0.5672408739725748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,5120,0.46935288111368817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,4096,0.02903466754489475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,3584,0.022658665974934895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,3584,0.33443644311692977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,3072,0.020190222395790946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,3072,0.2888000011444092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,2560,0.014159111513031853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,2048,0.008117333054542542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,2560,0.24316355917188856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,2048,0.19702488846249053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,1536,0.007052444749408298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,1536,0.15112711323632136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,1024,0.005384888913896348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,1024,0.10174755255381267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,768,0.004842666702138053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,768,0.07200355662239923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,512,0.00398577791121271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,512,0.047664887375301786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,256,0.0034844444857703317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,256,0.02808444367514716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,128,0.003149333306484752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,128,0.019295111298561096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,64,0.0029297777348094513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,64,0.016181333197487723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8192,32,32,0.003080000066094928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8192,32,32,0.014815110299322339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,12288,6.7419645521375875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,16384,8.342247856987848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,10240,6.646964603000217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,12288,8.116243150499132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,12288,5.540348476833767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,16384,10.710171169704862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,16384,7.456468370225694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8192,128,128,0.033864001433054604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,8192,3.699498494466146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,10240,5.563934326171875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,8192,5.332136789957683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,7168,3.2244302961561413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,7168,4.70994144015842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,10240,4.711907704671224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,8192,3.901929643419054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,6144,2.7826675838894315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,6144,4.0868712531195746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,5120,2.3074453141954208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,6144,2.891273710462782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,5120,3.5743200514051647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,4096,1.8417377471923828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,4096,3.1708560519748263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,5120,2.385296927558051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,3584,1.6211937798394098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,3584,2.389693366156684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,4096,1.9563973744710286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,3584,1.679085307651096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,3072,1.3885244793362086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,7168,7.1805614895290795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,3072,2.046455171373155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,2560,1.7096915774875219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,3072,1.3436275058322484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,2560,1.151158226860894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,2048,0.920146624247233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,2560,1.1903004116482205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,1024,0.7013697624206543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,2048,1.5005982716878254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,1536,1.0040551291571724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,1024,0.5154142379760742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,1536,0.6828248765733508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,2048,2.7505715688069663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,768,0.41249956025017637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,512,0.28026755650838214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,1536,0.7074444558885363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,256,0.24691200256347656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,1024,0.4564631250169542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,768,0.5459511015150282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,256,0.2683191034528944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,128,0.22426401244269478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,128,0.25478045145670575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,768,0.3495039939880371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,512,0.4293351173400879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,65536,512,0.40441687901814777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,256,0.27304087744818795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,128,0.23101422521803114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,64,0.20875022146436903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,64,0.22395377688937715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,65536,32,0.20323733488718668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,65536,32,0.21354399787055123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,16384,2.885207917955187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,16384,2.340663062201606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,12288,2.133079104953342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,12288,1.8515324062771266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,16384,2.4621476067437067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,10240,1.693078253004286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,12288,1.8534649742974176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,10240,1.5395164489746094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,10240,1.5262809329562719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,8192,1.1766470803154838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,8192,1.410903082953559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,7168,1.2800203959147136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,8192,1.2419164445665147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,65536,11.017478095160591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,7168,1.0492062038845487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,6144,0.909072028266059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,5120,0.8605519930521647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,6144,1.0259564717610676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,7168,1.0648790995279949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,6144,0.9365377426147461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,5120,0.755657778845893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,5120,0.7745466762118869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,4096,0.6701999770270454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,4096,0.6269484625922309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,3584,0.5316258006625706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,4096,0.6055973370869955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,65536,9.875678168402777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,3584,0.6445991198221842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,65536,16.549129909939236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,3072,0.537458684709337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,3584,0.5574053128560384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,3072,0.45757246017456055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,2560,0.428821325302124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,2048,0.3029866748385959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,1536,0.26548888948228627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,2048,0.7892968919542102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,3072,0.4768826696607802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,2560,0.3837457762824164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,2560,0.4141342110104031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,2048,0.33657066027323407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,1536,0.22299554612901476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,1024,0.18858577145470512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,1536,0.26256712277730304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,1024,0.15130221843719482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,1024,0.1904026667277018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,768,0.14614310529496935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,768,0.11766578091515435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,768,0.1546880006790161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,512,0.10677955547968547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,512,0.0936008890469869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,256,0.06953866614235772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,64,0.0554328891966078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,512,0.1185964478386773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,256,0.07626933521694608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,128,0.0596604413456387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,256,0.09045600228839451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,128,0.06835911009046766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,16384,128,0.08449066347546047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,64,0.06924266947640313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,16384,32,0.05576444334454007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,16384,32,0.06636889113320245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,16384,2.130185869004991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,12288,1.5209324094984267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,16384,1.932369867960612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,12288,1.5039022233751085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,10240,1.2342284520467122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,10240,1.2751102447509766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,12288,1.5530559751722548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,8192,0.9556648466322156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,16384,3.9574258592393665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,10240,1.2844070858425565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,8192,0.9794702529907227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,7168,0.812763532002767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,65536,7.820210774739583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,6144,0.6965324613783094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,8192,1.0356719758775499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,7168,1.133546617296007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,7168,0.9142026901245117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,6144,0.7547369003295898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,65536,8.881537543402777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,5120,0.567570686340332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,4096,0.5262444284227159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,5120,0.6267200046115452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,3584,0.45810757742987734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,6144,0.7766364415486654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,4096,0.5029519928826226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,5120,1.420634693569607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,3584,0.4442088868882921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,4096,0.5367137591044108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,65536,8.289466010199652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,3072,0.3758666780259874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,3584,0.4728079901801215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,3072,0.3848906622992621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,3072,0.4056897693210178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,2560,0.3188133239746094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,2560,1.4292898178100586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,2048,0.2571635511186388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,2560,0.35333598984612363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,2048,0.25424623489379883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,2048,0.28712444835238987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,768,0.10915733046001858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,1024,0.16547021600935194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,1536,0.19103999932607016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,1536,0.18518843915727404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,1536,0.22193600071801078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,1024,0.20602133538987902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,1024,0.12494222323099773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,768,0.09966578086217244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,256,0.05795555644565158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,768,0.13134577539232042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,256,0.07808355490366618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,512,0.08364444308810765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,512,0.07807466718885633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,512,0.10247466961542766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,256,0.06395022074381511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,128,0.04865600003136528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,128,0.055758224593268506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,64,0.042840000655916005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,12288,128,0.0724088880750868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,64,0.056999113824632436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,12288,32,0.043622222211625844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,12288,32,0.0551146666208903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,16384,1.8259752061631944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,16384,1.7176897260877821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,12288,1.168958240085178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,12288,1.3292889065212674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,16384,1.851666768391927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,10240,1.046937730577257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,10240,1.1176426145765517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,12288,1.4036222034030492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,8192,0.8874488936530219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,8192,0.8720782068040637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,10240,1.1673111385769315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,7168,0.7860888904995389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,65536,6.683778550889756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,7168,0.7795048819647895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,6144,0.6256782213846842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,8192,0.9285973442925347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,6144,0.676821337805854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,7168,0.8241324424743652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,5120,0.562554677327474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,5120,0.5000141991509331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,6144,0.706775983174642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,4096,0.41074665387471515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,5120,0.5978870921664768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,4096,0.451033804151747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,3584,0.3478719923231337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,3584,0.4336951043870714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,4096,0.4883848826090495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,3584,0.39596088727315265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,65536,7.468659294976129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,3072,0.33205689324273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,3072,0.34162312083774143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,2048,0.2216026650534736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,2560,0.28112978405422634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,3072,0.3741626739501953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,2560,0.28584088219536674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,2560,0.3188390996721056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,2048,0.22798487875196669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,1024,0.11194666226704915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,2048,0.2636551062266032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,1536,0.17014311419592965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,1536,0.16638488239712185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,1024,0.1185048951043023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,1536,0.20305066638522676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,65536,14.888209025065104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,768,0.09489511118994819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,1024,0.1458239952723185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,768,0.08931822246975368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,768,0.1208462185329861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,512,0.07226755883958605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,512,0.07011110915078057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,512,0.0938355525334676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,64,0.03616088959905837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,128,0.06527911292182074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,256,0.05242044395870633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,256,0.05605866511662801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,10240,256,0.07115199830796984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,128,0.03951199849446615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,128,0.04916533496644762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,64,0.04915111263593038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,10240,32,0.03681955403751797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,10240,32,0.049230221245023936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,16384,1.228798230489095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,16384,1.5125830968221028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,12288,0.9902266396416558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,12288,1.1658924950493705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,16384,1.6573306189643011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,10240,0.8730355368720161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,12288,1.2488524119059246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,10240,0.9831280178493924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,8192,0.6396951145595974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,10240,1.0444559521145291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,8192,0.7755422062344022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,7168,0.5441777971055772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,7168,0.688103993733724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,8192,0.8523813353644477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,6144,0.4910435676574707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,65536,6.659886254204644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,6144,0.5951866573757595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,5120,0.40775110986497665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,7168,0.7450773451063367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,6144,0.6416124237908257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,5120,0.496155579884847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,4096,0.3461715645260281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,5120,0.5344568888346354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,4096,0.4001564449734158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,65536,6.585309346516927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,3584,0.2846897708045112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,4096,0.44207199414571124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,3584,0.3500506612989638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,3072,0.2501315540737576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,3584,0.39149777094523114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,3072,0.3037431240081787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,2560,0.2175439993540446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,2560,0.2921875582800971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,3072,0.3411697811550564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,2560,0.25261155764261883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,2048,0.17162666055891249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,2048,0.20480088392893472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,1024,0.09892711374494766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,2048,0.23692801263597277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,1536,0.13480177190568712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,1536,0.14621244536505804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,1024,0.09527911080254449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,768,0.10908889108233982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,1536,0.18327732880910239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,1024,0.131059553888109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,768,0.07633777459462483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,768,0.07901422182718913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,256,0.04892622099982368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,512,0.05744622151056925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,512,0.06224444177415636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,65536,17.326831393771702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,256,0.03746933407253689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,512,0.08306133084826998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,128,0.03240888979699876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,256,0.06405599912007649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,128,0.04240977764129639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,64,0.029234667619069416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,8192,128,0.05753778086768257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,64,0.04284266630808512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,8192,32,0.030497776137457952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,8192,32,0.04186311033036974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,12288,0.7700737847222222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,16384,1.214735984802246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,12288,1.0862809287177193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,16384,1.4142666922675238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,10240,0.6535609033372667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,10240,0.9188666873508029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,8192,0.5496577686733669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,12288,1.1875279744466145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,10240,0.9980657365587023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,16384,2.4246258205837674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,7168,0.48060088687472874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,8192,0.722774240705702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,65536,4.498193952772352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,8192,0.8568853272332085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,7168,0.6358906428019205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,6144,0.4276248878902859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,6144,0.5556222067938911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,6144,0.6073964436848959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,5120,0.35706133312649196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,7168,0.995300398932563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,4096,0.28579288058810765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,5120,0.4619599978129069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,4096,0.3731253412034776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,3584,0.4817253218756781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,65536,5.903207143147786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,5120,0.5160897572835287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,3072,0.20965154965718588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,3584,0.32684622870551217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,4096,0.42332977718777126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,3584,0.37430310249328613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,3072,0.3528826766543918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,2560,0.19261333677503797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,3072,0.3193288909064399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,2560,0.23757865693834093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,2048,0.15858844916025797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,2560,0.2751413451300727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,2048,0.18900622261895073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,1024,0.0847253335846795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,1536,0.11732889546288384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,2048,0.29685865508185494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,1536,0.13690933916303846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,1536,0.17225866847568086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,1024,0.09354222483105129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,65536,6.284295823838975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,768,0.06825511323081122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,1024,0.12611022260453966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,768,0.07496355639563666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,512,0.0506586664252811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,512,0.05839733282725016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,768,0.11938844786749946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,256,0.036293334431118436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,512,0.07981599701775445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,256,0.04520888792143928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,128,0.029115554359224107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,256,0.061298668384552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,128,0.039036442836125694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,64,0.02970400121476915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,7168,128,0.05398933423890007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,64,0.0399217771159278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,7168,32,0.02751644452412923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,7168,32,0.03915822174814012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,12288,0.7452995512220594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,16384,1.31749513414171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,16384,3.099071078830295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,16384,1.4752186669243708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,12288,1.0045582453409831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,65536,3.887121836344401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,12288,1.108218616909451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,8192,0.4493875503540039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,10240,0.8451849089728461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,7168,0.41532444953918457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,10240,0.9293395148383247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,8192,0.6708960003323026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,7168,0.5916470951504177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,10240,2.1929403940836587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,8192,0.7556844287448459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,6144,0.3512729008992513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,65536,5.472301483154297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,6144,0.5134213235643175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,6144,0.5693608919779459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,7168,0.6694791052076551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,5120,0.42734487851460773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,4096,0.24722221162584093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,5120,1.4054906633165147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,3584,0.30164000723097056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,5120,0.48623376422458225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,4096,0.34417955080668133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,3584,0.22157866424984404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,4096,0.39581600824991864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,65536,5.851786719428168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,3072,0.20939200454288057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,3072,0.26163021723429364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,3584,0.35338756773206925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,2048,0.19264800018734404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,2560,0.15986043877071804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,3072,0.30520889494154185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,2560,0.21998133924272326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,2048,0.13562044832441542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,2560,0.2600826687282986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,1536,0.10408000151316325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,2048,0.21156267325083414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,1536,0.1661591132481893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,768,0.05884977844026354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,768,0.0695271094640096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,1536,0.1637608872519599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,1024,0.07328799698087904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,256,0.02958400050799052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,1024,0.08648266394933064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,256,0.04184533490075005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,768,0.09660889042748345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,1024,0.11955732769436306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,512,0.04584266742070516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,512,0.054712888267305165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,512,0.07613777452045016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,64,0.03696533375316196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,128,0.02546577817863888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,256,0.05819911426968045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,128,0.03586666782697042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,64,0.02415200074513753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,6144,128,0.05193688803248935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,6144,32,0.025648888614442613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,6144,32,0.03591288791762458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,16384,0.808686203426785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,12288,0.5711866484747993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,16384,1.2095911237928603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,12288,0.9247599707709419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,10240,0.5411653518676758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,16384,1.3838329315185547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,8192,0.40854843457539874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,12288,1.033765368991428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,10240,0.7760648727416992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,10240,0.877071115705702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,8192,0.6177333196004232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,7168,0.3441609011756049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,7168,0.5436862309773763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,6144,0.31350400712754994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,8192,0.6998471154106988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,6144,0.47451824612087673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,7168,0.6281590991550022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,5120,0.2702711158328586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,6144,0.534373336368137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,5120,0.3934648831685384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,4096,0.22372266981336805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,4096,0.3193662166595459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,5120,0.4562426673041449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,3584,0.18757332695855033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,65536,4.888078901502821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,3072,0.1620746718512641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,4096,0.37079999181959367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,3584,0.28056444062127006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,3072,0.24339734183417427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,3584,0.3324551052517361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,2560,0.13151821825239393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,2560,0.2026897801293267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,2048,0.1097066667344835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,3072,0.282641781700982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,2560,0.2473991182115343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,2048,0.1626124382019043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,1536,0.0853688915570577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,1536,0.1179342269897461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,2048,0.2004631095462375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,1024,0.06422933604982164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,65536,5.481109195285374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,1536,0.15416267183091906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,1024,0.08017155859205458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,768,0.05174044436878628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,768,0.06454133325152926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,1024,0.1118497716055976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,512,0.039719998836517334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,512,0.05067200130886502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,768,0.09404889080259536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,256,0.027772444817754958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,512,0.0720942219098409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,128,0.0326186650329166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,256,0.03827555643187629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,128,0.022521777285469904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,256,0.05454755491680569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,64,0.01978577839003669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,32,0.0324248903327518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,5120,128,0.04818222257826063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,5120,64,0.033237334754731916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,65536,8.990873548719618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,5120,32,0.020029332902696397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,16384,0.6065760188632542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,12288,0.46580176883273655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,16384,1.100054211086697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,12288,0.8404729101392957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,10240,0.39380976888868546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,12288,0.9903973473442925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,8192,0.31615376472473145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,10240,0.7051724327935113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,65536,2.6039271884494357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,10240,0.8183671103583442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,8192,0.5578142272101508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,7168,0.26878311898973256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,16384,3.129529741075304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,7168,0.48873244391547305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,8192,0.6685466766357422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,6144,0.25177955627441406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,6144,0.4278426700168186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,7168,0.591619544559055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,5120,0.19995555612776014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,65536,4.421872033013238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,5120,0.35363290044996476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,4096,0.15604621834225124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,4096,0.28664710786607533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,5120,0.4302684466044108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,4096,0.34679555892944336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,3584,0.15207110510932073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,6144,0.8518995708889432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,3584,0.24811999003092447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,3072,0.1273395617802938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,3072,0.26621688736809623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,3584,0.3076666726006402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,3072,0.21214132838779023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,2560,0.11231999927096897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,2560,0.19763644536336264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,1536,0.07174133592181735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,2048,0.08993689219156902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,2048,0.1402204434076945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,1536,0.16569777329762778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,1024,0.05019288924005297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,2560,0.22920799255371094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,65536,5.098889668782552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,2048,0.18788622485266793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,1536,0.0999564462237888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,1024,0.0704560014936659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,768,0.05468444360627068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,1024,0.10229244497087266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,768,0.05763644642300076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,768,0.08613777822918361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,512,0.03245333168241713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,512,0.04628444380230374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,128,0.028805332051383123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,512,0.06704622507095337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,64,0.014642667439248828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,256,0.022231111923853557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,256,0.034279998805787824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,128,0.01660355594423082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,256,0.05110399921735128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,64,0.028794666131337483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,4096,32,0.015409777561823526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,4096,128,0.04428799947102865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,4096,32,0.028412444723976985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,16384,0.6358453432718912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,12288,0.47665956285264754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,16384,1.0517572826809354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,12288,0.8018070856730143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,10240,0.3956009017096625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,8192,0.31467199325561523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,10240,0.6743377579583062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,65536,2.6085599263509116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,7168,0.2760995493994819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,8192,0.5339377721150717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,7168,0.4709751341078017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,6144,0.22990756564670137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,6144,0.4081244468688965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,65536,6.280641767713759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,5120,0.1923795541127523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,5120,0.4411199887593587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,16384,1.2918595208062065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,4096,0.15582666132185194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,12288,0.9814728630913628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,10240,0.8230515586005317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,4096,0.2746231026119656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,3584,0.1368808878792657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,3584,0.2399511072370741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,3072,0.12425243854522705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,7168,0.5901386472913954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,8192,0.6679502063327365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,3072,0.20639467239379883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,2560,0.10316533512539333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,2560,0.17332000202602812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,2048,0.08255733384026422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,2048,0.13712444570329455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,6144,0.5106710857815212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,65536,5.144972483317058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,1536,0.0666524436738756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,1536,0.09969244400660197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,1024,0.0487022234333886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,1024,0.06873066557778253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,5120,0.5178986655341254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,768,0.05414399835798475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,768,0.0667440030309889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,3584,0.3128257857428657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,512,0.02979466650221083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,2560,0.23104622628953722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,512,0.04382310973273384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,256,0.020261334048377145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,256,0.03221599923239814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,4096,0.35341776741875547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,1536,0.1436720026863946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,128,0.015061333775520325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,3072,0.2704177697499593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,128,0.027293332748942908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,64,0.01426488823360867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,64,0.02681955529583825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3584,32,0.01536711057027181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,2048,0.18719822830624053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3584,32,0.026485333840052288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,65536,1.9788684844970703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,16384,0.4974355697631836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,768,0.08672533432642619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,1024,0.10639199945661758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,65536,3.9813351101345487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,16384,1.0035119586520724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,12288,0.3787679937150743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,16384,1.519032902187771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,12288,0.763754685719808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,10240,0.32100534439086914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,65536,4.7449624803331165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,10240,0.66357421875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,8192,0.2506399949391683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,12288,0.9140026304456922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,8192,0.5097075568305122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,10240,0.7585466702779134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,7168,0.22538932164510092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,8192,0.6915315522087945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,7168,0.4494604534573025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,6144,0.1847440004348755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,7168,0.5463573137919108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,6144,0.38979199197557235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,5120,0.15725955698225233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,6144,0.6230382389492458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,5120,0.32453423076205784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,4096,0.12406578328874375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,5120,0.39896355734931094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,4096,0.2639484405517578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,3584,0.11489421791500515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,4096,0.5115448633829752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,3584,0.23043378194173178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,3072,0.09539555841022068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,3072,0.3100444475809733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,2560,0.1670560042063395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,3584,0.2890595595041911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,3072,0.19873155487908256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,2560,0.08660178052054511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,2048,0.06998311148749457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,2560,0.3207884364657932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,2048,0.1318204402923584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,1536,0.053674668073654175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,2048,0.17424710591634116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,1024,0.06582044230567084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,1536,0.09544710980521308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,1024,0.039035555389192365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,1536,0.16120889451768664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,768,0.03299199872546726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,768,0.052946666876475014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,1024,0.09657688935597737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,512,0.02901600135697259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,768,0.08022844129138522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,512,0.04184000028504265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,256,0.05205066667662727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,512,0.06333244509167142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,256,0.01790577835506863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,512,0.0685102211104499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,256,0.03205510973930359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,256,0.04820800158712599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,128,0.013354666531085968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,128,0.024915556112925213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,64,0.012095111111799875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3072,128,0.03901066713862949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,64,0.02442311081621382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,3072,32,0.012871999707486896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,3072,32,0.023574221465322707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,16384,0.3887084325154622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,65536,1.661217795477973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,12288,0.2963013384077284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,10240,0.2402613427903917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,12288,0.7315182156032987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,16384,1.886502159966363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,10240,0.6144026650322808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,8192,0.7532551023695203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,8192,0.4858204523722331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,65536,3.787807888454861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,7168,0.18027377128601074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,7168,0.4264666769239638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,6144,0.15416799651251897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,6144,0.3707919915517171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,12288,0.9164809121025933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,16384,1.2136578030056424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,10240,0.7668035295274523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,5120,0.1346613301171197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,5120,0.308114661110772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,4096,0.10988444752163357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,4096,0.24932622909545898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,8192,0.8689182069566516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,3584,0.09730755620532566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,3584,0.25133244196573895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,7168,0.5507004525926378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,3072,0.08854933579762776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,65536,4.722945743136935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,3072,0.1888871060477363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,2560,0.07048266463809542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,2560,0.1591386662589179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,2048,0.057222220632765025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,6144,0.47617599699232316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,2048,0.12629244062635633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,1536,0.04437244600719876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,1536,0.09048000309202407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,3584,128,0.044328000810411244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,1024,0.03135111265712314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,1024,0.06264266702863905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,3584,0.2901431189643012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,768,0.02589955594804552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,768,0.05075733198059929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,5120,0.39911378754509824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,2560,0.21485155158572725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,512,0.019884443945354886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,512,0.039763556586371526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,4096,0.32829509841071236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,256,0.014115555418862237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,256,0.02685333291689555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,128,0.010431999961535135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,128,0.022065778573354084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,1536,0.13262932830386692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,3072,0.252766211827596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,64,0.009019555317031013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,64,0.021697777840826247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,2048,0.17325245009528267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2560,32,0.009310222334331935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2560,32,0.021151110529899597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,65536,1.3065048853556316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,16384,0.31908265749613446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,16384,0.903166241115994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,65536,3.5908105638292103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,12288,0.23863466580708823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,16384,1.1181057824028862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,12288,0.6900515556335449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,10240,0.19953510496351454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,12288,0.8458657794528537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,65536,6.800697750515408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,10240,0.5813048680623373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,10240,0.708530690934923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,8192,0.4950879944695367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,8192,0.46271467208862305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,7168,0.5421857833862305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,8192,0.5725199911329482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,7168,0.4066142241160075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,6144,0.12533333566453722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,6144,0.35150755776299375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,5120,0.10301422410541111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,6144,0.43806666798061794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,7168,1.7650293774074977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,5120,0.2942435476515028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,4096,0.10067466894785564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,5120,0.36774399545457626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,4096,0.23799644576178658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,3584,0.07382844554053412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,3584,0.20903644296858045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,4096,1.0791386498345268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,3072,0.06581155459086101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,3584,0.267959992090861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,3072,0.35272977087232804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,2560,0.05386044581731161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,3072,0.2326915529039171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,2560,0.2234497865041097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,2560,0.1984471082687378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,2048,0.0446995562977261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,256,0.045687112543318004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,2048,0.1421066655053033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,128,0.03802311089303758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,2048,0.1609075599246555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,1536,0.03532444437344869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,1536,0.10202844275368585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,1536,0.12288978364732529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,1024,0.025459556115998164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,512,0.06374044550789727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,1024,0.05843377775616116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,768,0.021006221572558086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,768,0.04831111099984911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,1024,0.10335822237862481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,768,0.07406666543748644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,512,0.01646933290693495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,512,0.03609688745604621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,512,0.05608266592025757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,256,0.011258666714032492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,256,0.028527110815048218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,128,0.008560000194443597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,256,0.03984533415900336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,768,0.08127199941211276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,128,0.020498666498396132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,64,0.007541333635648091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2048,128,0.033724443780051336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,64,0.02103999919361538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,2048,32,0.00795999997191959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,2048,32,0.01958400011062622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,65536,1.0404684278700087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,16384,0.23668000433180067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,16384,0.8614844216240777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,12288,0.18622933493720162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,2560,1024,0.0979786647690667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,12288,0.6532764434814453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,65536,3.3928061591254344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,10240,0.2079119947221544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,10240,0.5498488744099935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,8192,0.1252382198969523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,8192,0.4391475518544515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,7168,0.11102755864461263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,7168,0.3865804407331679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,6144,0.09707644250657822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,6144,0.33404355578952366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,5120,0.0793057746357388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,5120,0.2807564470503065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,16384,1.0660933388604057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,4096,0.06797333558400472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,12288,0.806023120880127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,4096,0.22611554463704428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,3584,0.056256890296936035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,3584,0.19989332887861463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,10240,0.6778311199612088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,65536,4.201104905870226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,3072,0.04946844445334541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,8192,0.5578133265177408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,6144,0.42004889912075466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,3072,0.17258932855394152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,2560,0.04376444551679823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,7168,0.48330222235785586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,2560,0.14509067270490858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,5120,0.3535902235243056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,2048,0.035775999228159584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,2048,0.1171991162829929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,1536,0.028001778655582007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,1536,0.08418311013115777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,1024,0.0207342223988639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,1024,0.055953780810038246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,768,0.01643288963370853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,768,0.04505155483881632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,3584,0.25763821601867676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,512,0.012652444342772165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,512,0.033257779147889875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,2560,0.1890195608139038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,4096,0.2899271117316352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,256,0.008987555901209513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,256,0.023023999399609033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,768,0.06975289185841878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,128,0.007152888509962294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,128,0.01919822229279412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,64,0.006342222293217977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,64,0.018382221460342407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1536,32,0.00655644428398874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1536,32,0.018087110585636564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,3072,0.22209244304233125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,65536,1.0122462378607857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,1536,0.11741866005791558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,16384,0.15787911415100098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,65536,3.1931324005126953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,2048,0.1535795529683431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,16384,0.8130195405748156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,12288,0.12099644872877334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,12288,0.6164586808946397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,16384,1.0248204337226021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,10240,0.10351288980907863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,12288,0.7723520067003039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,8192,0.08376799689398871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,10240,0.5168719821506077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,10240,0.6526684231228298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,8192,0.4163653320736355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,7168,0.07456355624728732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,1024,0.0844586690266927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,8192,0.5279244316948785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,6144,0.06393422020806207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,7168,0.36635200182596844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,65536,5.610399034288194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,7168,0.4672702153523763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,6144,0.3158026536305745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,5120,0.06717511018117268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,6144,0.4052097797393799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,4096,0.046922667158974536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,5120,0.2662435637580024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,5120,0.33985333972507054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,4096,0.2157528930240207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,3584,0.039854221873813205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,4096,0.40017244550916886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,3584,0.18974133332570395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,3072,0.034565332863065935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,3584,0.24741599294874403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,3072,0.17166577445136177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,2560,0.029581334855821397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,3072,0.21360354953342012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,2560,0.138300445344713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,2048,0.024680000212457445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,1536,0.019488000207477145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,2048,0.11143822140163845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,2560,0.18984356191423205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,1024,0.014327110515700446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,1536,0.08012533187866211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,2048,0.14812444316016302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,1536,0.1132062276204427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,1024,0.055120001236597695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,768,0.011933333343929715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,1024,0.08193244536717732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,768,0.04248977700869242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,512,0.009389333426952362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,768,0.0680382251739502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,512,0.03134311238924662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,256,0.007098666495747036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,128,0.005948444621430502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,512,0.0513991117477417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,256,0.022333333889643352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,256,0.0373911095990075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,128,0.01777866648303138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1024,128,0.03134222163094415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,64,0.0053004444473319584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,32,0.016507556041081745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,1024,64,0.017474666237831116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,1024,32,0.00563466673096021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,65536,0.48032087749905056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,16384,0.12352266576555039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,16384,0.8021519978841146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,12288,0.12070577674441868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,16384,1.00200621287028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,12288,0.6083146731058756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,10240,0.08077777756585015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,65536,3.152800030178494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,10240,0.5098862118191189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,8192,0.0637706650627984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,12288,1.4965332878960504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,10240,0.6371582349141439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,8192,0.41041334470113117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,7168,0.13021955225202772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,8192,0.5160924593607584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,6144,0.0888320008913676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,128,0.031974222924974226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,7168,0.3605315685272217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,65536,3.9058125813802085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,6144,0.5076106919182671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,7168,0.4571386443244086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,6144,0.3113635645972358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,5120,0.04111822115050422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,5120,0.2622702121734619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,4096,0.06236355834537082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,4096,0.21295555432637533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,5120,0.332893345091078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,256,0.03841955463091532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,3584,0.03085244364208645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,4096,0.27165865898132324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,3584,0.18810044394599068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,3584,0.26330222023857963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,3072,0.026163554853863184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,3072,0.162463108698527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,2560,0.022728888524903193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,3072,0.20808000034756133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,2560,0.1362017790476481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,2048,0.020610666937298246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,2560,0.17802843782636854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,2048,0.11037600040435791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,1536,0.015283554792404175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,2048,0.14443466398451063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,1536,0.07950577470991346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,1536,0.10894578033023411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,1024,0.01204711116022534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,1536,512,0.05301600032382541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,1024,0.05251111255751716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,768,0.009544000029563904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,1024,0.08002755377027723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,768,0.04160177707672119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,512,0.007688889073001013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,768,0.06583733028835721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,512,0.03154933452606201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,256,0.005976888868543837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,512,0.04900799857245552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,256,0.021557332740889654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,128,0.005072888814740711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,256,0.036473777559068464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,128,0.017791999710930716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,768,128,0.030585779084099665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,64,0.004565333326657613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,64,0.016723554995324876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,768,32,0.004771555463473002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,768,32,0.01644177734851837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,65536,0.3365395598941379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,16384,0.08754399749967788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,12288,0.06652000215318468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,10240,0.05471377902560764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,16384,0.7612595558166504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,12288,0.5758559968736436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,8192,0.046560890144772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,10240,0.4839830928378635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,7168,0.03968355390760634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,8192,0.38979289266798234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,7168,0.3439733452267117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,6144,0.03395288851526048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,6144,0.2975573274824354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,65536,2.9874729580349393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,5120,0.030720889568328857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,5120,0.25055911805894643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,16384,0.9856772952609592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,4096,0.02476355599032508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,10240,0.6286328633626302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,8192,0.5076657931009928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,12288,0.7512124379475912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,3584,0.022273777259720698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,4096,0.2033004495832655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,3072,0.01867822143766615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,3584,0.17955288622114393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,7168,0.4473484357198079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,3072,0.15521244208017984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,2560,0.0163111107216941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,2560,0.13060088952382407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,2048,0.013244444297419654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,2048,0.10507555802663167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,6144,0.3883182207743327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,1536,0.010905777414639791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,1536,0.0744311081038581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,5120,0.3281671206156413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,1024,0.008462222086058723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,1024,0.04993333419164022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,65536,5.206806182861328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,768,0.01148444414138794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,768,0.04935200015703837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,3584,0.23599288198682997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,512,0.006114666660626729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,4096,0.2675662305619982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,512,0.028382221857706707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,3072,0.20565422375996908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,256,0.005014222115278244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,2560,0.17361156145731607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,256,0.020095111595259774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,128,0.004399111287461387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,128,0.01645777788427141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,64,0.004048000193304486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,64,0.015282667345470853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,512,32,0.004193777839342753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,2048,0.14137510458628336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,512,32,0.015054222610261707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,65536,0.2098471058739556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,1536,0.1068871087498135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,16384,0.0614853302637736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,1024,0.0783546699417962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,768,0.06549866994222005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,12288,0.052263110876083374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,16384,0.7519546614752876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,12288,0.6065502166748047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,10240,0.05394311083687676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,65536,2.948996437920464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,10240,0.5194995668199327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,8192,0.03295466634962294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,8192,0.4044915570153131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,128,0.02944177720281813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,7168,0.031065778599845037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,512,0.04822577701674568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,512,256,0.035240888595581055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,6144,0.028100444210900202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,7168,0.3397066593170166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,6144,0.2938444349500868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,5120,0.02513155506716834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,5120,0.2474506696065267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,16384,0.9714462492201063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,4096,0.019004444281260174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,4096,0.20097155041164824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,12288,0.7351635297139486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,3584,0.017059556312031217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,3584,0.17704444461398652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,10240,0.6208080185784234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,3072,0.015623110863897534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,65536,3.8041119045681424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,8192,0.5232204331292046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,3072,0.15353777673509386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,2560,0.013173333472675748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,2560,0.1291724443435669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,2048,0.011019555230935415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,2048,0.10388533274332683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,7168,0.4402773380279541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,6144,0.3816826608445909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,1536,0.008971555365456475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,1536,0.07379377550548978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,1024,0.006991110742092133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,1024,0.04909066690338982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,5120,0.32163556416829425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,768,0.005954666684071223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,768,0.03956444395913018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,4096,0.2626124487982856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,3584,0.23220356305440268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,512,0.005015111217896144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,512,0.02773422168360816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,256,0.004049777984619141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,256,0.019744000501102872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,3072,0.20149600505828857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,2560,0.1717440022362603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,128,0.0036311112344264984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,128,0.01609244445959727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,64,0.0033902221669753394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,64,0.014888000157144336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,256,32,0.0035111109415690103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,256,32,0.014627556006113688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,65536,0.1944213310877482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,2048,0.1395288838280572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,65536,2.932272805107964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,16384,0.0649964412053426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,16384,0.7492639753553602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,1536,0.1054346693886651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,768,0.06345155504014757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,12288,0.0388408899307251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,1024,0.07646844122144911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,10240,0.0333262218369378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,12288,0.5668444633483887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,10240,0.4749164581298828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,8192,0.026534222894244727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,8192,0.38376177681816953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,256,0.03391199972894456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,7168,0.023934221929974023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,128,0.028664890262815688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,256,512,0.047021332714292735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,7168,0.3383919927808974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,6144,0.02179733415444692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,6144,0.29277155134412974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,5120,0.017027555240525138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,5120,0.2467191219329834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,16384,0.9656231138441297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,4096,0.01034755590889189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,4096,0.20000355773501924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,65536,4.432002597384982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,3584,0.011647111011876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,3584,0.17921245098114014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,12288,0.7315973175896539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,3072,0.00870577742656072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,10240,0.6141111056009928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,3072,0.15292799472808838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,2560,0.007528000407748752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,2560,0.12856088744269475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,8192,0.4963795344034831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,2048,0.0064186664919058485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,2048,0.10271644592285156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,7168,0.43882134225633407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,1536,0.005717333406209946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,6144,0.3791804313659668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,1536,0.07351911067962646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,1024,0.00479911102188958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,1024,0.04870666729079353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,5120,0.32003821267022026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,768,0.004298666699065102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,768,0.039008001486460366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,4096,0.26055555873446995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,512,0.003850666599141227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,512,0.027468444572554693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,3584,0.2307048903571235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,256,0.003439999909864532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,256,0.019464888506465487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,3072,0.20066932837168375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,128,0.0031715554909573663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,128,0.015626667274369132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,2560,0.1700942251417372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,64,0.0030497778207063675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,64,0.014479999740918478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,128,32,0.0032924444725116095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,128,32,0.014202665951516895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,2048,0.13746578163570827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,65536,0.1844142278035482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,16384,0.05125066637992859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,1536,0.10392710897657607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,12288,0.041165331999460854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,16384,0.7480079862806531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,10240,0.03515911102294922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,12288,0.5666133562723795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,768,0.062092443307240806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,8192,0.029306666718588933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,1024,0.07514133056004842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,8192,0.38441067271762425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,10240,0.4750995635986328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,7168,0.0266275554895401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,6144,0.02124622298611535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,5120,0.015360888507631091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,7168,0.3384088940090603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,65536,2.9368125067816844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,6144,0.29294223255581325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,4096,0.008786666724416945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,5120,0.24684532483418783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,3584,0.008003555238246918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,4096,0.19997778203752306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,3072,0.008496888809733922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,3584,0.1767022212346395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,2560,0.007635555333561367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,3072,0.1522995498445299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,2560,0.12835644351111516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,2048,0.006923555499977536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,1536,0.005520888914664586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,2048,0.10265422529644436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,1536,0.07345777750015259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,1024,0.004578666554556953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,768,0.004113777644104428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,1024,0.04857422245873345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,256,0.003277333246337043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,512,0.027681777874628704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,512,0.003681777666012446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,768,0.03905955619282193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,256,0.01959822244114346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,128,0.003056888985964987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,128,0.015785778562227886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,64,0.002875555513633622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,64,32,0.002963555562827322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,64,0.014532445205582513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,64,32,0.014105778601434497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,65536,0.1794639958275689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,16384,0.049786667029062905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,12288,0.03961866762903001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,10240,0.03346666693687439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,16384,0.750333309173584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,512,0.04751111070315043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,12288,0.5669573148091634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,8192,0.028063111835055884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,10240,0.4757511350843642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,7168,0.025416889124446448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,8192,0.3845573266347249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,6144,0.02275644408331977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,7168,0.33858667479621035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,5120,0.014514666464593677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,6144,0.29260089662339955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,4096,0.007920000288221572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,5120,0.24704355663723418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,256,0.03378488951259189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,3584,0.007424888511498769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,65536,2.9404044681125217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,3072,0.15286133024427626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,4096,0.2004257837931315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,2048,0.005438222239414851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,3584,0.17621066835191515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,3072,0.0068488890926043195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,2560,0.006050666587220297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,2560,0.12824622790018717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,2048,0.10327111350165473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,1536,0.004791111167934206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,1536,0.07333778010474311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,1024,0.0041422223051389056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,256,0.003121777748068174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,768,0.003773333297835456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,1024,0.04879111051559448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,768,0.03887644410133362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,512,0.003495111233658261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,64,0.014451555079884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,512,0.027355555031034682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,128,0.0028817777832349143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,256,0.01945955554644267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,128,0.015767110718621146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,64,0.0027582223216692605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4096,32,32,0.0028844444702068963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4096,32,32,0.014145778285132514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,12288,2.762844509548611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,16384,3.5066320631239147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,10240,3.3604950375027127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,16384,5.30748536851671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,10240,2.3028106689453125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,12288,6.854118347167969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,8192,2.7001156277126737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,8192,1.8243893517388239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,7168,1.601186646355523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,6144,2.016480975680881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,7168,5.342600928412543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,6144,1.3815413581000433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,5120,1.818380355834961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,5120,1.142563607957628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,12288,2.8006969028049045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,16384,3.8000318739149304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,4096,1.4059697257147894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,10240,2.3455164167616105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,4096,0.9190933439466688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,3584,0.7949244181315104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,8192,1.751548449198405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,3584,1.2138071060180664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,7168,1.5504817962646484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,3072,1.0152613321940105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,3072,0.6793422169155545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,2560,0.8238746854994031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,2560,0.5565235349867079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,2048,0.6447706752353245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,2048,0.44318487909105087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4096,128,128,0.028128001425001357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,1536,0.5053093168470595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,6144,5.1300739712185335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,5120,1.0885439978705513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,1536,0.3325200080871582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,1024,0.2249680095248752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,1024,0.39252532853020566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,768,0.29841865433586967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,768,0.18812089496188691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,3584,0.7877786424424914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,512,0.22654488351609972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,512,0.1544640064239502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,4096,0.8754293123881022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,256,0.14647289117177328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,256,0.13264532883961996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,2560,0.5688879754808214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,128,0.11755288971794976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,128,0.12278132968478733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,3072,0.6745368639628092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,64,0.11647555563184951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,64,0.12219022379981147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,65536,32,0.11248000462849934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,65536,32,0.11755377716488308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,1536,0.543756432003445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,768,0.22915556695726183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,2048,0.4720640182495117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,16384,1.3466836081610785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,65536,5.150746663411458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,16384,1.1632711622450087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,12288,0.9746515485975477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,12288,0.8960435655381945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,16384,1.2188630633884006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,1024,0.2746044529808892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,65536,6.396305084228516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,10240,0.8347182273864746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,10240,0.753209802839491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,8192,0.6353359752231175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,12288,0.9133902655707465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,8192,0.5946106910705566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,7168,0.5737013287014431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,7168,0.5186515384250218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,8192,0.6207360161675347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,10240,1.3940088484022353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,65536,4.916375054253472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,6144,0.4436222182379828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,6144,0.4547457695007324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,7168,0.5478240119086372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,5120,0.4320026768578424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,5120,0.3755493428972032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,4096,0.3373395601908366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,6144,0.7477679782443576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,5120,0.3997511068979899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,4096,0.3036026689741346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,3584,0.2751679950290256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,4096,0.3168506622314453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,3584,0.4332088894314236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,3072,0.25367556677924263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,3072,0.21910756164126924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,3584,0.28665243254767525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,2560,0.17953866057925752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,3072,0.240685330496894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,2048,0.18337778250376383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,2560,0.21079199843936494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,2048,0.14316888650258383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,1536,0.13347644276089138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,2560,1.0139840443929036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,1536,0.1092275579770406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,2048,0.17257600360446504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,1024,0.09699289004007976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,1536,0.14191999700334337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,1024,0.18309777312808564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,768,0.0675191084543864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,768,0.07693155606587727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,1024,0.10777778095669216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,512,0.09630222453011407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,768,0.09168089098400539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,512,0.055204444461398654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,512,0.07274933656056722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,256,0.038125332858827375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,256,0.06194577614466349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,128,0.033005333609051175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,64,0.032609777318106756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,128,0.14266755845811632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,256,0.059769776132371694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,128,0.04205422268973457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,16384,128,0.05743822124269274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,64,0.04404000110096402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,16384,32,0.03358488943841722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,256,0.14044177532196045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,16384,32,0.04264800084961785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,65536,512,0.17341421710120308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,16384,0.9950461917453342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,12288,0.6756693522135416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,12288,0.7371084425184461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,10240,0.5768515798780653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,16384,1.018807093302409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,16384,3.262517293294271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,12288,0.7696257697211372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,10240,0.6214284367031521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,65536,3.894415113661024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,8192,0.4895022180345323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,10240,0.6469013426038954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,65536,3.945436265733507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,7168,0.4108106560177273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,8192,0.5291777716742622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,7168,0.4285253418816461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,6144,0.36668888727823895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,6144,0.37340354919433594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,5120,0.2989822228749593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,5120,0.3085751003689236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,8192,2.1487645043267145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,7168,0.4665030903286404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,4096,0.22865600056118437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,6144,0.3947689003414578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,5120,0.34033600489298504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,4096,0.2482302188873291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,3584,0.21501155694325766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,4096,0.2739351060655382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,3072,0.19140799840291342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,3072,0.18024088276757133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,3584,0.9726248847113715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,3584,0.2382462289598253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,2560,0.15361777941385904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,3072,0.2099235587649875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,2560,0.14913511276245117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,2048,0.12867289119296604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,2560,0.17951377232869467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,1536,0.0994044476085239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,2048,0.18870755036671957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,1536,0.0912968913714091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,2048,0.14814399348364934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,1024,0.09383910894393921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,768,0.08914489216274685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,1024,0.07251022259394328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,1536,0.12314844131469727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,1024,0.06788000133302477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,65536,4.133432176378038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,768,0.05761510796017117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,512,0.04450133442878723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,768,0.08001422219806247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,512,0.06832444667816162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,128,0.03521777855025397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,256,0.02970044480429755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,512,0.06483733654022217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,64,0.036680890454186335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,256,0.03871377640300327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,128,0.025474665893448725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,256,0.05708355373806424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,64,0.02544266647762722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,12288,128,0.05011733373006185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,12288,32,0.0255084451701906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,12288,32,0.03596355517705282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,16384,0.7830808957417806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,12288,0.6306542290581597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,12288,0.6547644403245714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,10240,0.4816524187723796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,16384,2.331360075208876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,16384,0.9264461729261609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,10240,0.5520488950941298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,12288,0.7071377966139051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,8192,0.4353875584072537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,7168,0.3483440081278483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,65536,3.335368898179796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,10240,0.5853350957234701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,8192,0.47437508900960285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,8192,1.4004364013671875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,7168,0.38309600618150497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,65536,3.4822567833794484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,6144,0.3097635640038384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,6144,0.3324897819095188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,5120,0.2753422260284424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,7168,0.4268035623762343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,4096,0.21111378404829237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,5120,0.6508853170606825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,5120,0.3033493359883626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,6144,0.3645511203342014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,4096,0.22782844967312285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,3584,0.1943075524436103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,3072,0.15652533372243246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,4096,0.25386044714185924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,3584,0.20007377200656465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,3072,0.16022311316596136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,2560,0.13869688245985243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,3072,0.1879031128353543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,2560,0.1316817733976576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,2048,0.1081848873032464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,2560,0.16469866699642605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,3584,0.4327244493696425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,2048,0.10605600145128037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,1536,0.08877333005269368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,2048,0.13907644483778211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,1536,0.08364178074730767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,1024,0.06376177734798856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,1536,0.11372444364759658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,1024,0.07749955521689521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,768,0.0514977773030599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,1024,0.08568177620569865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,65536,3.7310167948404946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,768,0.05245511068238152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,512,0.039000001218583845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,768,0.07424533367156982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,512,0.04459555612670051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,256,0.05022666520542569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,256,0.028219554159376357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,512,0.06021866533491346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,256,0.0380968881977929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,128,0.02206666602028741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,128,0.0315048893292745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,64,0.019938665959570143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,10240,128,0.04816710948944092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,64,0.03312355611059401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,10240,32,0.020114666885799833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,10240,32,0.03198666705025567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,12288,0.4714648988511827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,16384,0.7504408624437121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,10240,0.4029875596364339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,16384,2.3602373335096574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,12288,0.580767101711697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,16384,0.837774223751492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,65536,2.582895067003038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,12288,0.635392877790663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,8192,0.3236622280544705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,10240,0.4906933572557237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,8192,0.38700177934434676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,7168,0.28918398751152885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,65536,3.0440453423394094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,7168,0.33762489424811465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,6144,0.23185955153571236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,8192,0.4404746691385905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,6144,0.29390043682522243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,7168,0.38213245073954266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,5120,0.2005751132965088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,5120,0.24509334564208984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,4096,0.164682666460673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,4096,0.19586666425069174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,6144,0.3266426722208659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,3584,0.14150577121310765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,10240,1.5893004735310872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,5120,0.2848835521274143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,3584,0.16933866341908774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,4096,0.23127910825941297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,3072,0.12138844860924615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,3584,0.20055554972754586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,2560,0.11681244108412002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,3072,0.141503996319241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,2560,0.10688444640901353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,2560,0.15053066942426893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,2048,0.08619466755125259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,3072,0.27241423394944936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,2048,0.09328889184527928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,1536,0.06818133592605591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,2048,0.12566844622294107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,1536,0.07363822062810262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,1024,0.05106755428844028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,1536,0.10296444760428534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,1024,0.06393955813513862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,65536,3.317555533515083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,768,0.04121955566936069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,1024,0.07848088608847724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,256,0.022425777382320825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,768,0.04714044597413805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,512,0.031712889671325684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,768,0.06869155830807157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,512,0.03838933176464505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,256,0.031187554200490315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,512,0.0576488905482822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,128,0.018238221605618794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,256,0.045928888850741915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,128,0.027954667806625366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,64,0.014630221658282809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,8192,128,0.04182755615976122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,64,0.02903644575013055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,8192,32,0.015448000695970325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,8192,32,0.02817866537306044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,16384,0.5473128954569498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,12288,0.4178906546698676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,16384,0.7021182378133138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,12288,0.7309510972764758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,10240,0.35483021206325954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,10240,0.45526398552788633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,8192,0.27380622757805717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,65536,2.205906762017144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,8192,0.35837509897020126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,7168,0.25564445389641655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,7168,0.315591123369005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,65536,2.8078496721055775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,6144,0.21267999543084037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,6144,0.2746346791585286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,5120,0.17974400520324707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,5120,0.2274613380432129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,16384,0.8484897613525391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,12288,0.6428986655341254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,4096,0.15047289265526667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,4096,0.18524621592627633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,10240,0.7932400173611112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,3584,0.21107910739050972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,7168,0.3880488872528076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,3584,0.1873475578096178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,8192,0.43953421380784774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,3072,0.1470408837000529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,3072,0.133041779200236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,6144,0.3378915521833632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,2560,0.08999199999703301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,2560,0.10904622077941895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,5120,0.28447643915812176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,2048,0.0734071135520935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,2048,0.08797422382566665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,1536,0.061121775044335254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,1536,0.06940355565812853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,1024,0.04050577680269877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,65536,3.30804083082411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,1024,0.05225422316127353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,768,0.03281866510709127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,768,0.04500444398985969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,3584,0.20488977432250977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,512,0.02517066730393304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,2560,0.1509066687689887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,512,0.03629600008328756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,256,0.017673777209387887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,256,0.029473778274324205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,768,0.06880533032947116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,128,0.01461955573823717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,1536,0.10359999868604873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,128,0.026305778159035578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,64,0.013570666313171387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,64,0.0266942216290368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,4096,0.23267555236816406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,7168,32,0.014930667148696052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,7168,32,0.02478222217824724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,3072,0.17697066730923125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,2048,0.12607555919223362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,65536,1.9667076534695094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,16384,0.44146132469177246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,16384,0.654679987165663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,65536,2.599160936143663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,12288,0.3546373314327664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,12288,0.4975004196166992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,16384,0.7526408831278483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,10240,0.3026204374101427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,1024,0.0813795526822408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,10240,0.42165067460801864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,12288,0.5704000261094835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,8192,0.24781955613030326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,8192,0.333934227625529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,7168,0.21138577991061738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,10240,0.48204178280300564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,8192,0.391175111134847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,7168,0.29240889019436306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,6144,0.18659733401404488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,6144,0.2539573245578342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,7168,0.34482754601372617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,5120,0.16423466470506456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,6144,0.30137332280476886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,5120,0.21210043960147432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,4096,0.1303457816441854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,5120,0.25284889009263783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,4096,0.1721280018488566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,3584,0.112225784195794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,4096,0.20628800657060412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,3584,0.14718577596876356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,3072,0.09751644399431016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,3584,0.18155554930369058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,3072,0.12222577465905084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,2560,0.08321244186825223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,3072,0.1560951073964437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,2560,0.10190577639473809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,2048,0.06792977783415052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,2048,0.08170666959550646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,2560,0.1344026724497477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,1536,0.05376266770892673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,2048,0.11336266994476318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,1536,0.0649048884709676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,1024,0.04011199871699015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,1024,0.04918577935960558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,1536,0.09247377845976089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,768,0.03236088818973965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,65536,6.125448015001084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,1024,0.07192889187071058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,768,0.04240266813172234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,768,0.06236888964970907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,512,0.05049777693218655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,512,0.02516977820131514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,256,0.018056000272432964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,512,0.05189866820971171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,256,0.027815110153622095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,256,0.048026667700873486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,128,0.01386844449573093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,128,0.02365066607793172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,64,0.026093333959579468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,64,0.012109333442317115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,6144,32,0.012429333395428128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,6144,128,0.03761777612898085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,6144,32,0.02296888828277588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,16384,0.3807013299730089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,16384,0.6041830910576714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,65536,2.395457797580295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,12288,0.30521954430474174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,12288,0.4599182340833876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,10240,0.252162668440077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,10240,0.3876817756228977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,65536,7.6985380384657125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,8192,0.9875191582573785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,128,0.042678223715888135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,8192,0.9907342063056098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,256,0.046762665112813316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,7168,0.27123822106255424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,7168,0.8679973284403483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,7168,512,0.05753778086768257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,6144,0.7476746771070691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,6144,0.23514310518900552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,5120,0.12486044565836589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,5120,0.19622311327192518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,16384,0.7539644241333008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,4096,0.10535289181603326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,4096,0.1577999989191691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,12288,0.5725679927402073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,3584,0.09569955534405178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,3584,0.13552710745069715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,65536,2.9237467447916665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,3072,0.08489244514041477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,10240,0.48159466849433047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,3072,0.1128871175977919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,2560,0.07045955790413751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,2560,0.09286489089330037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,8192,0.39499110645718044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,2048,0.059226665231916636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,2048,0.0757804446750217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,7168,0.34664445453219944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,1536,0.043997334109412305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,1536,0.060439109802246094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,6144,0.4133324358198378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,1024,0.06384533643722534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,1024,0.0808782246377733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,5120,0.25366756651136607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,768,0.02609599961174859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,768,0.039260443713929914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,3584,0.18365689118703207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,512,0.02016888890001509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,512,0.03148622314135233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,2560,0.13620000415378147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,256,0.014290667242474027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,256,0.024875554773542616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,4096,0.2060755491256714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,128,0.010785777535703448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,128,0.021038222644064162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,3072,0.15695644749535456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,64,0.009152889251708984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,64,0.021576000584496394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,5120,32,0.009514666265911525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,5120,32,0.02082577844460805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,65536,1.335826661851671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,768,0.06243289179272122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,1536,0.09307555357615153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,16384,0.29888444476657444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,65536,2.1697154574924045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,2048,0.11422577169206406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,16384,0.5515519777933756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,12288,0.22600889205932617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,16384,0.6527262263827854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,12288,0.4229884412553575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,10240,0.19858133792877197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,12288,0.5025395817226833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,10240,0.35663734541998965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,8192,0.15673777792188856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,10240,0.42428000768025714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,8192,0.2835608853234185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,7168,0.14265955819023982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,8192,0.3469582133822971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,7168,0.24992267290751138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,65536,2.5480738745795355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,6144,0.12348000208536784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,7168,0.3049857881334093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,6144,0.2169546683629354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,5120,0.10063911146587795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,5120,0.18075022432539198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,4096,0.0800879995028178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,6144,0.26424090067545575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,4096,0.1452017757627699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,3584,0.07528266641828749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,5120,0.22561955451965332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,4096,0.18450221750471327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,1024,0.07382311423619588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,3584,0.12406399514940049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,3072,0.06484355529149373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,3584,0.16065333949195013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,3072,0.10375466611650254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,2560,0.05327111151483324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,3072,0.13998044861687556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,2560,0.0849306649631924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,2048,0.04500888784726461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,2560,0.1177626715766059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,2048,0.06980178091261122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,2048,0.10734933614730835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,1536,0.035383111900753446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,1536,0.05587110916773478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,1536,0.0831786659028795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,1024,0.026132444540659588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,768,0.021159110797776118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,1024,0.040469331873787776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,1024,0.061448891957600914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,768,0.03651911020278931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,768,0.05552000138494703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,512,0.028941333293914795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,512,0.01661333276165856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,256,0.011633777783976661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,256,0.021925333473417494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,512,0.04511555698182848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,128,0.008781332936551835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,128,0.019202666150199044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,64,0.00757244477669398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,256,0.035953776703940496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,4096,32,0.008023111356629265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,4096,128,0.033859554264280535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,64,0.01949155496226417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,4096,32,0.019229332605997723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,16384,0.2978275616963704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,16384,0.5474106470743815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,65536,1.1838231616550021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,12288,0.2344426578945584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,16384,0.6649004618326823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,12288,0.41955288251241046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,10240,0.19226845105489096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,12288,0.5051315625508627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,10240,0.3534826702541775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,8192,0.1555075512992011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,65536,2.582700517442491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,8192,0.2814088927374946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,10240,0.42713154686821836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,7168,0.18121600151062012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,7168,0.24751999643113878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,8192,0.3483582337697347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,6144,0.1466791099972195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,65536,2.5615225897894964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,6144,0.21450400352478027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,6144,0.26651732126871747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,7168,0.3096880118052165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,5120,0.1199680036968655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,5120,0.1792551146613227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,4096,0.08125688632329305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,5120,0.22554755210876465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,4096,0.14398133754730225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,128,0.03740533192952474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,3584,0.16197421815660265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,3584,0.07135288582907783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,256,0.0413991113503774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,4096,0.1879333390129937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,3584,0.12345688872867161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,3072,0.06299377812279595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,3072,0.10310933325025771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,3072,0.1387066708670722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,2560,0.053582221269607544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,2560,0.08523822493023342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,2048,0.0447599987188975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,1536,0.05560533205668131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,5120,512,0.05149955550829569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,2560,0.11997600396474202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,2048,0.06810755862130059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,768,0.02107911143038008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,1024,0.0637635588645935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,1536,0.03529777791765001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,2048,0.09989421897464329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,512,0.028212444649802312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,1024,0.025968000292778015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,1536,0.08254488971498278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,1024,0.04055022199948629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,768,0.03575199842453003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,512,0.016214221715927124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,768,0.05706044700410631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,256,0.011058666639857821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,512,0.046479112572140165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,256,0.021736888421906367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,128,0.008471999731328752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,256,0.03635377685228983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,128,0.018984888990720112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,64,0.007207999626795451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3584,128,0.03339111142688327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,64,0.019175999694400363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3584,32,0.007878222399287755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3584,32,0.018966222802797954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,65536,0.9554782443576388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,16384,0.7049999766879611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,12288,0.19026133749220106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,16384,0.5045590930514866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,12288,0.3866737683614095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,10240,0.2354177898830838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,10240,0.3261982334984673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,8192,0.12530755996704102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,65536,1.9790577358669703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,8192,0.2608239915635851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,7168,0.11115288734436035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,7168,0.22845511966281465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,6144,0.09461244609620835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,6144,0.19742133882310656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,5120,0.08193333281411065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,5120,0.16517066955566406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,16384,0.6131102244059244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,8192,0.3208906650543213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,4096,0.06806844472885132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,4096,0.13412977589501276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,3584,0.05812266800138685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,12288,0.4665573438008626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,3584,0.11579467190636529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,10240,0.3926533328162299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,3072,0.049435555934906006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,65536,2.3646702236599393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,3072,0.09621066517300075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,2560,0.042100446091757886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,2560,0.07892355653974745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,2048,0.03439111179775662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,2048,0.06287644306818645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,7168,0.29701511065165204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,6144,0.24615022871229383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,1536,0.0369466675652398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,1536,0.05037066671583387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,1024,0.01998933321899838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,1024,0.03763822383350796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,5120,0.20918667316436768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,768,0.01681422193845113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,768,0.03237333231502109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,3584,0.1492124398549398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,512,0.013257777525318993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,512,0.025255999631351892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,2560,0.11150844891866048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,768,0.05061155557632446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,256,0.009305777649084726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,256,0.019959110352728102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,128,0.0071937781241205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,128,0.017829333742459614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,4096,0.17042488521999785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,1536,0.07677066988415189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,64,0.00628977765639623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,64,0.017987555927700467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,3072,32,0.006535111202134027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,3072,32,0.017818666166729398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,65536,0.8055377536349826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,3072,0.12870044178432888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,16384,0.19572888480292427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,16384,0.49927377700805664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,65536,2.0189395480685763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,12288,0.24927200211418998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,2048,0.09294222460852729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,12288,0.3958711094326443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,10240,0.16433689329359266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,10240,0.3563857873280843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,1024,0.058561775419447154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,8192,0.10278311040666367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,8192,0.2557555569542779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,128,0.031579554080963135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,7168,0.09046133359273274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,7168,0.2236888938479953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,256,0.033977776765823364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,6144,0.07673333088556926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,6144,0.1956577830844455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,3072,512,0.04360266526540121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,5120,0.0651600029733446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,5120,0.16380533907148573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,12288,0.45312976837158203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,4096,0.05338577760590447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,4096,0.13192889425489637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,16384,0.5971813201904297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,65536,2.2987217373318143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,10240,0.4058364497290717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,3584,0.048136889934539795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,3584,0.11400889025794135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,3072,0.042656001117494374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,3072,0.0947164429558648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,2560,0.03512622250450982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,2560,0.07811199956470065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,8192,0.3126257790459527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,2048,0.030228445927302044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,2048,0.06267555554707845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,7168,0.2771066559685601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,1536,0.024007999234729346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,1536,0.05003822180959913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,6144,0.2405306763119168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,1024,0.017504889104101393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,1024,0.037500444385740496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,5120,0.2044044468137953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,768,0.014799111419253878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,768,0.031585776143603854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,3584,0.14696978198157415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,512,0.011676444775528379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,512,0.025063999825053748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,2560,0.10779733128017849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,256,0.00832622249921163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,256,0.01984444426165687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,4096,0.16592621803283691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,128,0.006786666810512543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,128,0.017846221725145977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,3072,0.12552800443437365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,64,0.006050666587220297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,64,0.017900443739361234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,2048,0.09132977988984849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2560,32,0.00629333323902554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2560,32,0.01773244473669264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,65536,0.5890026622348362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,65536,1.801655027601454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,16384,0.1591928932401869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,16384,0.4587164454989963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,16384,0.5756604406568739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,12288,0.12077511681450738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,65536,2.1960603925916886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,12288,0.352310233645969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,10240,0.10149155722724067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,12288,0.4371315638224284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,10240,0.29709866311815053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,8192,0.08235822121302287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,10240,0.36760889159308535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,8192,0.23669332928127715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,7168,0.07168533404668172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,8192,0.3013342221577962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,7168,0.20804533693525526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,6144,0.06170666880077786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,7168,0.26643021901448566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,6144,0.18027022149827746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,5120,0.05221688747406006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,6144,0.23169244660271537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,5120,0.15087555514441595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,4096,0.04249777727656894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,5120,0.19563021924760607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,4096,0.12174222204420303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,3584,0.03756622142261929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,4096,0.1588782204522027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,3584,0.10427733262379964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,3072,0.03360889024204678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,3584,0.13946221934424505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,768,0.05019911130269369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,3072,0.08685421943664551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,2560,0.02870222263866001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,3072,0.12006489435831706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,2560,0.07155111100938585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,2048,0.024471999870406255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,2560,0.10311555862426758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,2048,0.057038221094343394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,1536,0.01956622302532196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,2048,0.0872382190492418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,1536,0.04521066612667508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,1024,0.014912888407707214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,1536,0.07176978058285184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,1024,0.03369777732425266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,768,0.012311110893885294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,1024,0.05505422088834974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,768,0.028083556228213843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,512,0.009763555394278632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,768,0.0459502206908332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,512,0.02237155536810557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,256,0.007174222005738153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,512,0.038860443565580584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,256,0.01815733313560486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,128,0.006039999839332368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,128,0.016270221935378183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,256,0.03202133377393087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,64,0.005391110976537068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2048,128,0.02995377779006958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,64,0.016393777396943834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,2048,32,0.005560889012283749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,2048,32,0.016178665889634024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,1536,0.07585066556930542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,65536,0.4433324337005615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,16384,0.12337422370910645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,16384,0.4572284486558702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,16384,0.5452355808681911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,12288,0.0895306666692098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,12288,0.3446417649586995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,65536,1.7645742628309462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,10240,0.07787022325727674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,12288,0.4159822199079726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,8192,0.06260799699359469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,10240,0.29130487971835667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,10240,0.35087378819783527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,65536,2.0964505937364364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,8192,0.23314666748046875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,7168,0.05584444602330526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,8192,0.2878053453233507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,6144,0.04891555507977804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,7168,0.20582489172617593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,6144,0.17767022715674508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,7168,0.2531191243065728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,5120,0.04061155517896017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,6144,0.22125421630011666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,4096,0.03417333298259311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,5120,0.149073772960239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,5120,0.18605511718326143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,4096,0.12021866109636094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,3584,0.029575109481811523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,4096,0.15267466174231634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,3584,0.10410488976372613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,3072,0.02609688871436649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,3072,0.0862648884455363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,2560,0.022498667240142822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,3584,0.1333635515636868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,3072,0.1137582196129693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,2560,0.0716115567419264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,1024,0.057441777653164335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,2048,0.019288889235920377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,2560,0.0995982223086887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,2048,0.05727466609742907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,1536,0.01568533314598931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,2048,0.08343644274605645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,1536,0.04569155640072293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,1024,0.01146666705608368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,1536,0.06939288642671373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,1024,0.03350755572319031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,768,0.009768889182143742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,1024,0.05319111214743721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,768,0.02956977817747328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,512,0.007847111258241866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,512,0.02276266614596049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,768,0.04778666628731621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,256,0.005978666659858491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,512,0.037547555234697126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,256,0.018409777018758986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,128,0.005062222066852781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,256,0.03120000163714091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,128,0.016501333978441026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,64,0.004583111239804162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,64,0.01628177695804172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1536,128,0.02921600143114726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1536,32,0.004842666702138053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,128,0.030417776770061914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1536,32,0.016193777322769165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,65536,0.3038871023390028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,16384,0.0868062244521247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,12288,0.06432622008853488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,16384,0.413191106584337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,256,0.03253600001335144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,12288,0.31621156798468697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,10240,0.05320533447795444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,10240,0.2665377722846137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,8192,0.0431253347131941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,65536,1.6009066899617512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,8192,0.21543378300136992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,7168,0.038703110482957624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,7168,0.1894684500164456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,6144,0.03327466713057624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,6144,0.1646417776743571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,2560,512,0.04157155421045091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,5120,0.030258665482203167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,5120,0.1382017797893948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,16384,0.5294604301452637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,4096,0.023919110496838886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,12288,0.40541956159803605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,4096,0.11111911137898763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,3584,0.020554666717847187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,10240,0.34098222520616317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,3584,0.096596446302202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,3072,0.01835911141501533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,8192,0.27704087893168133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,3072,0.08038310872183906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,65536,2.041621314154731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,2560,0.015824000040690105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,7168,0.24540445539686415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,2560,0.06534488995869954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,2048,0.01371022231049008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,2048,0.05314844515588549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,1536,0.010961777634090848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,6144,0.21275822321573892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,1536,0.04228355487187704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,1024,0.00851822230550978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,1024,0.03122755553987291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,5120,0.18151377307044136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,768,0.007194666398896112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,768,0.026847110854254827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,3584,0.12825244002872044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,512,0.006072888771692912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,512,0.020796444680955675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,4096,0.1475528875986735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,256,0.005004444470008214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,256,0.017068444026841056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,2560,0.09516444471147324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,128,0.004406222038798862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,128,0.015125332607163323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,3072,0.11075466871261597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,2048,0.08067288663652208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,64,0.003916444463862313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,64,0.014880888991885714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,1024,32,0.004187555362780889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,1024,32,0.014615999327765571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,1536,0.06768088870578341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,65536,0.2501280042860243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,16384,0.06835555367999606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,1024,0.05082577798101637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,16384,0.40608710712856716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,12288,0.06017600165473091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,65536,1.5786497328016493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,12288,0.3094693289862739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,10240,0.04916622241338094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,10240,0.2615591155158149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,768,0.045200887653562755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,8192,0.03672444489267137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,8192,0.21024533112843832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,512,0.036079999473359846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,128,0.028770665327707928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,1024,256,0.03028977910677592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,7168,0.03183733423550924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,7168,0.1861120065053304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,6144,0.030810667408837214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,6144,0.16130044725206164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,5120,0.025384889708624944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,5120,0.13564088609483507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,12288,0.4022560119628906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,4096,0.020097777247428894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,16384,0.5276542239718968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,4096,0.10947999689314102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,3584,0.0169813334941864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,10240,0.3411591053009033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,3584,0.09372266795900132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,65536,2.0211493174235025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,3072,0.016857778032620747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,3072,0.07762933439678617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,2560,0.013352000051074557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,2560,0.06408266888724433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,8192,0.27681512302822536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,2048,0.012076444096035428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,2048,0.05170933405558268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,7168,0.24441422356499565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,1536,0.009088888764381409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,1536,0.04154400030771891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,6144,0.21228088272942436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,1024,0.007052444749408298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,1024,0.030226667722066242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,5120,0.17987110879686144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,768,0.006305777778228124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,768,0.02563200063175625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,3584,0.12886667251586914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,512,0.005351999981535806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,2560,0.0949324435657925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,512,0.020862221717834473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,256,0.004429333325889376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,256,0.016915554801623028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,4096,0.1469457811779446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,128,0.003973333372010125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,3072,0.11034666829639012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,128,0.014911999305089315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,64,0.003657777690225177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,64,0.014696889453464083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,768,32,0.003839111162556542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,768,32,0.014516444669829475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,65536,0.16527911027272543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,65536,1.5460017522176106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,2048,0.08056533336639404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,16384,0.05322844452328152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,16384,0.40121955341762966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,768,0.043495999442206494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,12288,0.048766222265031606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,12288,0.3053777747684055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,1536,0.06736533509360419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,10240,0.04312088754442003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,10240,0.25815823343065053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,1024,0.05070844292640686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,8192,0.02979555394914415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,8192,0.20930400159623888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,256,0.029873778422673542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,7168,0.02942133280966017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,7168,0.1847377750608656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,128,0.028117332193586562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,6144,0.025915554828113977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,6144,0.1589004463619656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,768,512,0.03561244408289591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,5120,0.02286044425434536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,5120,0.13495822747548422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,65536,1.974178738064236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,16384,0.5136088795132107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,4096,0.01714666684468587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,4096,0.1083822250366211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,3584,0.01555377741654714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,3584,0.09329066673914592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,12288,0.3932577768961589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,3072,0.014319999350441827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,3072,0.07712088690863715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,10240,0.3333751095665826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,2560,0.012135110795497894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,2560,0.06381689177619086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,8192,0.26994845602247447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,2048,0.010418666733635796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,2048,0.05131733417510986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,7168,0.23896710077921549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,1536,0.00828888929552502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,1536,0.04116977916823493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,6144,0.20729688803354898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,1024,0.006324444380071428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,1024,0.02975822157329983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,5120,0.17641511228349474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,768,0.005503111001518037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,768,0.025253333979182776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,4096,0.1434897714191013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,512,0.0047528888616296975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,512,0.020448888341585796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,3584,0.12488444646199544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,256,0.003954666770166821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,256,0.016562667157914903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,3072,0.10779822534985012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,2560,0.09306222200393677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,128,0.003593777616818746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,128,0.014818666709793938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,64,0.0033884443756606844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,64,0.014662222729788886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,512,32,0.003489777859714296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,512,32,0.014426666829321118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,65536,0.11267644829220241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,65536,1.5460248523288305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,2048,0.0785822206073337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,16384,0.03703644540574815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,16384,0.401789320839776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,1536,0.06613244613011678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,12288,0.030729777283138696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,12288,0.3064302338494195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,1024,0.049908445941077344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,10240,0.026916444301605225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,10240,0.2578071223364936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,768,0.042502221133973866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,8192,0.02053333322207133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,8192,0.20936266581217447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,256,0.029044446018007066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,7168,0.018106665876176622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,512,0.03472355670399136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,7168,0.18447466691335043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,6144,0.015767999821239047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,6144,0.15926488240559897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,512,128,0.027113777067926195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,5120,0.013944000005722046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,5120,0.13367556201087102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,65536,1.9493039449055989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,4096,0.011328000161382886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,4096,0.10751199722290039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,16384,0.5074799855550131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,3584,0.00962577760219574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,3584,0.09279821978674994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,12288,0.38779645495944554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,3072,0.009047110875447592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,3072,0.07704444726308186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,10240,0.3281431198120117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,2560,0.007516444557242923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,2560,0.06296444601482816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,8192,0.26700978808932835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,2048,0.0064382221963670515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,2048,0.050973332590527005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,7168,0.23530488544040254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,1536,0.00562755556570159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,1536,0.04085155659251743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,6144,0.20434310701158312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,1024,0.0047306666771570844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,1024,0.028528001573350694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,5120,0.17308621936374238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,768,0.004328888737493091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,768,0.023952888117896184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,4096,0.1413608921898736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,512,0.003938666648334927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,512,0.020195555355813768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,3584,0.12385600143008763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,256,0.003481777591837777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,256,0.016195555528004963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,3072,0.10675999853346084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,128,0.0032142222755485107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,128,0.014492443866199918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,2560,0.09116444322797988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,64,0.003027555636233754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,64,0.014158222410413953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,256,32,0.003175111073586676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,256,32,0.013901333014170328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,65536,0.10344355636172825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,2048,0.07765866650475396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,65536,1.546977784898546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,1536,0.06337866518232557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,16384,0.03161244591077169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,16384,0.40059465832180446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,12288,0.02516266703605652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,1024,0.04823733369509379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,12288,0.305552879969279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,10240,0.022301332818137273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,10240,0.2574515607621935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,768,0.041490667396121554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,8192,0.013762666947311826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,8192,0.20837688446044922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,512,0.03413866625891791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,7168,0.013001778059535556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,7168,0.18450400564405653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,256,0.028689778513378564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,6144,0.015066666735543145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,256,128,0.027365333504146997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,6144,0.15962489446004233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,5120,0.013944000005722046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,5120,0.13437333371904162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,16384,0.49909067153930664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,65536,1.9180062611897786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,4096,0.010635555618339114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,4096,0.10793777969148423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,3584,0.009672000176376766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,12288,0.380495998594496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,3584,0.09309333562850952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,3072,0.009118222528033787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,10240,0.32098576757642955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,3072,0.07668799824184842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,2560,0.007063999772071838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,2560,0.06260177824232313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,8192,0.26151466369628906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,2048,0.006309333360857434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,2048,0.05095555716090732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,7168,0.23211023542616102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,1536,0.005494222044944763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,1536,0.04088711076312595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,6144,0.20109689235687256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,1024,0.004629333400064045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,1024,0.02905333373281691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,5120,0.1708079973856608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,768,0.004134222037262387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,768,0.024534222152498033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,4096,0.13866578208075628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,512,0.003685333248641756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,512,0.01982311076588101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,3584,0.12086844444274902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,256,0.0032559999575217566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,256,0.016044444508022733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,3072,0.10505955749087864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,128,0.003072000005178981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,128,0.014349333114094205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,2560,0.08899111217922634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,64,0.0029235554652081597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,64,0.014093332820468478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,128,32,0.002987555538614591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,128,32,0.013845333622561561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,65536,0.09807733032438491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,2048,0.07547911008199056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,16384,0.029174221886528864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,65536,1.54694398244222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,16384,0.4010311232672797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,12288,0.025391111771265667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,10240,0.01956977778010898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,12288,0.30518754323323566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,8192,0.01181333346499337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,10240,0.2573404312133789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,7168,0.010441777606805166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,8192,0.20839288499620226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,7168,0.18432266182369658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,6144,0.015224888920783997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,6144,0.15866488880581325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,5120,0.008934222161769867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,5120,0.13376800219217935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,4096,0.008184889124499427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,4096,0.10762400097317165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,3584,0.007375111182530721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,3072,0.006800888727108638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,3584,0.09293244282404582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,3072,0.07636266946792603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,2560,0.006383111079533895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,2560,0.061871111392974854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,2048,0.00554044461912579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,2048,0.05121155579884847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,1536,0.00489155575633049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,1536,0.04010577665434943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,1024,0.0041404445138242514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,1536,0.06269688738716973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,1024,0.02789333462715149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,768,0.003830222205983268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,768,0.024149333437283833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,512,0.003504000190231535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,512,0.019905777441130746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,256,0.0031840000301599503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,256,0.016007110476493835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,128,0.0029404444826973807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,128,0.01442399952146742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,64,0.0027724444452259275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,64,0.014057776994175382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,1024,0.04724533359209696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,64,32,0.0028408887899584244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,64,32,0.013726222018400827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,65536,0.09467289182874893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,16384,0.028187556399239436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,768,0.0399235553211636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,12288,0.022647110952271357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,16384,0.4012133280436198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,10240,0.017044444878896076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,12288,0.3062160015106201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,8192,0.01016088906261656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,10240,0.25747821066114635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,7168,0.009742221898502773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,8192,0.20887111292945015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,65536,1.5508142047458227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,7168,0.18388711081610787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,6144,0.008749333520730337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,6144,0.15911289056142172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,5120,0.008218666745556725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,5120,0.13394667042626276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,4096,0.007426666716734569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,3584,0.007152000235186682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,4096,0.1074391139878167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,3584,0.09192711114883423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,3072,0.00675644435816341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,3072,0.07699910799662273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,512,0.03350399931271871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,2560,0.006093333164850871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,2560,0.06223999791675144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,2048,0.005369777894682354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,2048,0.05087111062473721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,1536,0.0047893333766195504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,256,0.028284443749321833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,1536,0.040066666073269315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,512,0.003479111111826367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,1024,0.0041697778635554844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,1024,0.029346668057971533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,768,0.0038844446341196695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,768,0.023775110642115276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,512,0.019837333096398246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,256,0.0031591111587153543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,256,0.01625688870747884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,128,0.002986666642957263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,128,0.014389332797792224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,64,0.0028213332924577924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,64,0.013925333817799887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2048,32,32,0.002899555489420891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2048,32,32,0.013774221969975365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,16384,1.8015545739067926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,12288,1.3747697406344945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,12288,2.032119115193685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,10240,1.7265244589911566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,16384,2.74983639187283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,10240,1.1491129133436415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,8192,0.9042533238728842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,8192,1.3763920466105144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,7168,1.1344711515638564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,7168,0.7864311006334094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,6144,0.6750337812635633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,6144,1.0445351070827908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,10240,1.1219511032104492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,8192,0.8956960042317709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,16384,1.8466328514946833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,7168,0.7899733119540744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,5120,0.5550390879313151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,5120,0.8620230886671277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,4096,0.7227768898010254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,4096,0.44988266626993817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,3584,0.6196622318691677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,3584,0.38791555828518337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,12288,3.4423092736138234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,6144,0.6485030916002061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,5120,0.5574960178799099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,3072,0.5179057651095921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,3072,0.33301512400309247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,4096,0.45116620593600804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,2560,0.2784995502895779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,2560,0.44573333528306747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,3584,0.40162134170532227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,2048,0.2208017773098416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,2048,0.3536070982615153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,2048,0.25275733735826283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,1536,0.16712978151109484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,2560,0.3017164336310493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,1536,0.2694151136610243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,3072,0.8742657767401801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,1024,0.1889653338326348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,1024,0.12019377284579807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,1536,0.2029351128472222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,1024,0.1533271074295044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,768,0.14973421891530356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,768,0.10131466388702393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,512,0.11245511637793647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,768,0.16848533683352998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,512,0.08638577991061741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,256,0.07814133167266846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,512,0.10487288898891872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,128,0.08662755621804132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,256,0.0750382211473253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,32,0.057690666781531445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,65536,256,0.0885288914044698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,32,0.07773155636257596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,128,0.06312177578608195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,128,0.06948710812462701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,65536,64,0.05799377626842923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,65536,64,0.07201777564154731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,16384,0.6025173399183485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,12288,0.46275467342800564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,16384,1.5875191158718531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,12288,0.4458115365770128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,65536,2.2985856797960067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,65536,2.6596177419026694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,10240,0.3738444381290012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,16384,0.6160551177130805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,12288,0.46560531192355686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,8192,0.33345866203308105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,10240,1.4797546598646376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,8192,0.2892684406704373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,7168,0.2691884570651584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,10240,0.3832124339209662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,7168,0.26021599769592285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,6144,0.23564177089267305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,8192,0.31895645459493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,6144,0.21425510777367485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,5120,0.17395022180345324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,5120,0.21532089180416533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,6144,0.23553244272867838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,4096,0.17116444640689424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,4096,0.1431235604816013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,5120,0.2061084508895874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,3584,0.14965244134267172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,7168,0.8225484424167209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,3584,0.12386133935716416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,4096,0.1703022188610501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,3072,0.12353777885437012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,3584,0.15425422456529406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,3072,0.1754773325390286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,2560,0.11109689209196304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,2560,0.09308089150322808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,3072,0.13679556051890054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,2048,0.11489778094821507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,2048,0.07496088743209839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,2560,0.11853688293033177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,1536,0.08270577589670818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,65536,2.478961732652452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,1536,0.08577511045667861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,1536,0.0706204440858629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,2048,0.10081777969996135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,1024,0.052100443177753024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,1024,0.06072000000211927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,768,0.04253422220547994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,1024,0.06852355268266466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,768,0.04053511222203573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,768,0.05938666396670871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,512,0.03267022305064731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,512,0.035353776481416486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,256,0.022984888818528917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,512,0.052894221411810986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,256,0.02972444560792711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,256,0.04432622260517544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,128,0.01774311065673828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,128,0.0273626661962933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,16384,128,0.04081777731577555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,64,0.01494311127397749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,64,0.028042667441897925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,16384,32,0.015627556376987033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,16384,32,0.029264890485339697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,16384,0.46748712327745223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,16384,0.5630906422932943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,12288,0.34544889132181805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,65536,1.9820747375488281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,12288,0.3621813456217448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,65536,1.8906160990397136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,16384,0.5259875721401638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,10240,0.3061590989430745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,10240,0.2886648972829183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,8192,0.2388311227162679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,8192,0.2374702294667562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,10240,0.33670843972100156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,7168,0.20246488518185088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,8192,0.27995644675360787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,7168,0.20618755287594268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,12288,1.144171502855089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,6144,0.1877253320482042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,6144,0.17431910832722983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,7168,0.24117689662509492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,5120,0.1484933296839396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,5120,0.14685155285729304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,6144,0.20663200484381783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,4096,0.1235742171605428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,5120,0.17625689506530762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,4096,0.11744622389475505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,65536,2.040122561984592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,3584,0.13806845082177058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,3072,0.08893955416149563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,3584,0.13109866778055826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,4096,0.14517066213819715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,3584,0.1035955548286438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,3072,0.09933600160810684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,3072,0.11797955301072861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,2560,0.0838586688041687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,1536,0.05443466703097025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,2560,0.07783999707963732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,2048,0.07038578059938219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,2560,0.10234133402506511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,2048,0.06502666738298205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,2048,0.08875733613967896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,1536,0.05268889003329807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,1024,0.03876266545719571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,1536,0.07573599947823419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,1024,0.04065955678621928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,1024,0.060152000851101346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,768,0.0329386658138699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,768,0.03565066721704271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,512,0.02539288832081689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,256,0.026011554731263056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,768,0.05573955509397718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,512,0.030743112166722614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,256,0.018028444714016385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,512,0.04630577895376417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,256,0.040072888135910034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,128,0.014006222287813822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,128,0.023590222001075745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,64,0.012269333004951477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,12288,128,0.03688177797529433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,64,0.022867555419603985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,12288,32,0.01254666679435306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2048,128,128,0.027087110612127516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,12288,32,0.02313866714636485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,16384,0.42406755023532444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,12288,0.28684888945685494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,65536,1.6273742251926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,16384,1.736196411980523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,65536,1.680031140645345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,12288,0.3244240019056532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,10240,0.23846401108635795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,10240,0.4194035530090332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,8192,0.19182133674621582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,8192,0.22331110636393228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,7168,0.17127377457088896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,7168,0.18364889091915557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,6144,0.14676177501678467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,6144,0.156896882587009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,5120,0.13403556081983778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,5120,0.12802933322058782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,16384,0.5255057546827528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,4096,0.11077688799964057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,12288,0.3990231090121799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,4096,0.1046560009320577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,10240,0.3355315526326497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,3584,0.09857600265079075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,3584,0.09291466739442612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,3072,0.08242044183943006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,8192,0.3606977727678087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,3072,0.08035111427307129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,7168,0.24237333403693306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,2560,0.07022488779491849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,2560,0.07044888867272271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,65536,2.0250256856282554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,2048,0.05805777841144138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,2048,0.058535999721950956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,6144,0.208369771639506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,1536,0.04609244399600559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,1536,0.04833955566088358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,1024,0.03304266598489549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,5120,0.17691644032796225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,1024,0.03666222095489502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,768,0.02699022160636054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,768,0.031823111904992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,3584,0.13333511352539062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,512,0.020584889584117465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,512,0.02794044547610813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,2560,0.10338399807612102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,256,0.014209777116775513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,256,0.024102222588327196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,4096,0.14601955148908827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,768,0.052503112289640635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,128,0.010865777730941772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,1536,0.0757697754436069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,128,0.02149244480662876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,64,0.009352000223265754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,64,0.020787555310461257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,10240,32,0.009404444032245213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,10240,32,0.02066044509410858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,3072,0.1181982225841946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,2048,0.08878222439024185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,65536,1.3048310809665256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,16384,0.30637065569559735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,65536,1.4871973461574977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,16384,0.37798489464653867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,12288,0.22288799285888672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,12288,0.2893093427022298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,16384,0.43195022477044004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,10240,0.21247555149926078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,12288,0.33157777786254883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,10240,0.24172798792521158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,8192,0.15130755636427137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,8192,0.18990666336483428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,10240,0.2812657886081272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,7168,0.14545422130160862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,7168,0.16356889406840006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,8192,0.22586133744981554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,6144,0.12754666805267334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,6144,0.13720622327592638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,7168,0.19993600580427381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,5120,0.09748266802893744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,5120,0.11299822065565322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,6144,0.17298221588134766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,4096,0.0838479995727539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,4096,0.09203022056155735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,5120,0.14520711368984646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,3584,0.0758462217119005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,4096,0.12265688843197292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,3584,0.08172088861465454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,65536,2.3552044762505426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,3072,0.06741511159472995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,1024,0.06077510780758328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,3072,0.07151555352740817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,3584,0.11156000031365289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,2560,0.057710223727756076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,2560,0.06303733587265015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,3072,0.12191200256347656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,2048,0.04558666547139486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,2560,0.08846488926145767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,2048,0.05173866616355049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,1536,0.03875644339455499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,2048,0.0741093357404073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,1536,0.04382044408056471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,1024,0.02605688903066847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,1536,0.06536444690492418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,1024,0.03351022137535943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,1024,0.05101066827774048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,768,0.021215111017227173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,768,0.029578665892283123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,512,0.016499555773205228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,768,0.04500266578462389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,512,0.025676444172859192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,256,0.011955555942323474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,512,0.040729777680502996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,256,0.022477333744366963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,128,0.009097778134875828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,256,0.03620355659061008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,128,0.01906577746073405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,64,0.00756888919406467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,8192,128,0.03255733185344272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,64,0.019317333896954853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,8192,32,0.00795288880666097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,8192,32,0.01900800069173177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,16384,0.30203376875983345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,16384,0.3723706669277615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,65536,1.4611209233601887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,12288,0.22614132033454049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,65536,1.753959125942654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,12288,0.28426043192545575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,16384,0.43414754337734646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,10240,0.3810746669769287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,10240,0.23699821366204155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,12288,0.3339244524637858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,8192,0.21236356099446616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,8192,0.18565689192877877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,10240,0.2800248993767632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,7168,0.14781422085232204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,7168,0.1606933275858561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,8192,0.22902843687269422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,6144,0.16754844453599718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,7168,0.20126399728986952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,6144,0.1371555593278673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,65536,1.6764319737752278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,5120,0.14849778016408285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,5120,0.1117946704228719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,6144,0.17387733194563124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,4096,0.08329866992102729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,5120,0.14705688423580593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,4096,0.1492568916744656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,3584,0.07353244225184123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,4096,0.12337156136830647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,3584,0.08046755525800917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,3072,0.08482666810353597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,3584,0.11204889085557725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,3072,0.07052089108361138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,2560,0.054564442899492055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,3072,0.0999164448844062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,2560,0.0706533326043023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,2048,0.04605422086185879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,2560,0.08826844559775458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,2048,0.05115555392371284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,1536,0.039368887742360435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,2048,0.07547199726104736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,1536,0.04349422123697069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,1024,0.02524799936347538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,1536,0.06526222493913439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,1024,0.035007112556033664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,768,0.020792000823550753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,1024,0.05282844437493218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,768,0.029300444655948218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,768,0.04660355713632372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,512,0.01739377776781718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,512,0.025232000483406916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,256,0.011666666302416058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,256,0.02144177754720052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,512,0.04225511021084256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,128,0.00847999999920527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,256,0.03839022252294753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,128,0.01868711080816057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,64,0.007246221933099959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,7168,128,0.032650666104422674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,64,0.018988443745507132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,7168,32,0.008184000021881526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,7168,32,0.018723555737071566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,128,0.038314667012956403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,16384,0.32829154862297905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,16384,0.840012444390191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,512,0.046374221642812095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,65536,1.0230337778727214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,12288,0.17045333650377062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,10240,256,0.040361778603659734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,10240,0.1772142251332601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,65536,1.287109375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,10240,0.21131022771199545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,8192,0.12625688976711696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,12288,0.5863359769185384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,8192,0.1666275527742174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,7168,0.10661422544055515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,6144,0.08857422404819065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,7168,0.2115644481447008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,6144,0.12208978335062663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,5120,0.08487911356820001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,5120,0.09878933429718018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,12288,0.2959120008680555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,8192,0.20477422078450522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,16384,0.386515564388699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,10240,0.24974934260050455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,4096,0.061829335159725614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,7168,0.1788720024956597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,4096,0.08022844129138522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,6144,0.17608977688683403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,3584,0.05506933397716946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,3584,0.07199644380145602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,65536,1.4751413133409288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,3072,0.06715733475155301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,3072,0.062301331096225314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,2560,0.044016887744267784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,2560,0.06163377894295586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,2048,0.03580800029966567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,5120,0.13183644082811144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,2048,0.047134220600128174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,1536,0.02791822287771437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,1536,0.03827111257447137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,1024,0.020149333609475028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,1024,0.030638221237394545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,768,0.016732444365819294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,768,0.026922666364245947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,768,0.04223555657598707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,512,0.013232000172138214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,512,0.022807111342748005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,3584,0.09947022464540269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,256,0.009450666606426239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,256,0.018738667170206707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,2560,0.07895644505818684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,128,0.007149332927332983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,128,0.01737422247727712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,1536,0.05717421902550591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,64,0.006299555715587404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,64,0.01773244473669264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,6144,32,0.006591111007663939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,3072,0.08848710854848225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,6144,32,0.01757333344883389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,65536,0.8622328970167371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,4096,0.10869599713219537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,65536,1.2593680487738717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,16384,0.19863288932376436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,16384,0.32120267550150555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,2048,0.06813244687186347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,12288,0.15034577581617567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,12288,0.24686932563781738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,1024,0.04727377825313144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,10240,0.1340293354458279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,10240,0.20683022340138754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,8192,0.09714666340086196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,8192,0.16237600644429526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,128,0.030440890126758154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,7168,0.08581333027945624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,7168,0.14040977425045437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,256,0.03362311257256402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,6144,0.0793386697769165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,6144,0.11818133460150824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,6144,512,0.03809422254562378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,5120,0.06931555271148682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,5120,0.0983733336130778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,12288,0.2827093336317274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,4096,0.05334666702482435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,65536,1.3868213229709203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,4096,0.07852177487479316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,3584,0.04850577645831638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,10240,0.23936266369289824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,16384,0.3684835433959961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,3584,0.06968888971540663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,3072,0.042358222934934825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,3072,0.06121955315272013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,2560,0.0362462235821618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,2560,0.054382221566306226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,8192,0.19524266984727648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,2048,0.029622223642137315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,2048,0.04526844289567736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,7168,0.17222400506337485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,1536,0.023992889457278784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,1536,0.03748888770739237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,6144,0.14895999431610107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,1024,0.018028444714016385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,1024,0.02996088729964362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,5120,0.1266684399710761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,768,0.01479466590616438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,768,0.026221333278550044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,3584,0.09647999869452582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,512,0.0120000003112687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,2560,0.07815555731455485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,512,0.022664889693260193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,256,0.00848444468445248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,256,0.019632889164818656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,4096,0.10644533236821492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,128,0.007105777661005656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,128,0.017496888836224873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,3072,0.0865031083424886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,2048,0.06632800234688653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,64,0.006236444330877728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,64,0.017635555730925668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,5120,32,0.00647555540005366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,5120,32,0.017480888300471835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,65536,0.6281635496351454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,65536,1.0928977330525715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,65536,1.2852587170071073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,16384,0.15162221590677896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,16384,0.28031910790337455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,12288,0.11423733499315049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,16384,0.34288443459404844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,12288,0.2159679995642768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,12288,0.2610035472446018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,10240,0.09628444247775608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,10240,0.18148533503214517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,768,0.04189688960711161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,8192,0.07809955543941922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,10240,0.2227226628197564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,8192,0.14250400331285265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,7168,0.06836444139480591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,8192,0.1796640025244819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,7168,0.12331555949317084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,6144,0.06023110946019491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,7168,0.15830755233764648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,6144,0.10365866952472264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,5120,0.05060799916585287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,6144,0.1354115539126926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,5120,0.08427821927600437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,4096,0.04260977771547106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,4096,0.06834844748179118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,5120,0.11649333106146918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,3584,0.039461334546407066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,4096,0.09656888908810085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,3584,0.061236441135406494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,3072,0.033180445432662964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,3584,0.08903288841247559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,3072,0.05446311169200473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,2560,0.029559999704360962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,3072,0.07912177509731717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,2048,0.0237137774626414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,2560,0.04811111092567444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,2560,0.07260088788138495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,2048,0.039820445908440485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,2048,0.06110933091905382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,1536,0.019419555862744648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,1536,0.033061332172817655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,1024,0.014450665977266101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,1536,0.051925334665510386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,1024,0.02647022240691715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,768,0.012335999972290464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,1536,0.05615999963548449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,1024,0.043360888957977295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,768,0.02283199959331089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,512,0.009926222264766693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,512,0.01941866676012675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,768,0.03944266504711575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,256,0.007121777368916406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,512,0.03579555617438422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,256,0.01689511040846507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,128,0.0058453331391016645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,128,0.016041777200169034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,256,0.030922667847739324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,64,0.005260444349712796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,4096,128,0.029121776421864826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,64,0.01613333324591319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,4096,32,0.015761777758598328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,4096,32,0.00566133318675889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,16384,0.1504817803700765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,16384,0.27504444122314453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,65536,1.2714915805392795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,65536,1.0730426576402452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,12288,0.13429689407348633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,12288,0.20960356129540336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,16384,0.34415555000305176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,10240,0.09575555721918742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,10240,0.17659910519917807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,12288,0.39172087775336367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,65536,1.295328034294976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,8192,0.08007999923494127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,8192,0.14033599694569907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,10240,0.22164977921379936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,7168,0.06955199771457247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,7168,0.1209191083908081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,8192,0.17988977167341444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,6144,0.059306667910681836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,7168,0.15779466099209255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,6144,0.10405333174599542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,1024,0.04654666781425476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,5120,0.05023733443684048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,5120,0.08290400107701619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,6144,0.13626044326358372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,4096,0.04013155566321479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,5120,0.1882399982876248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,4096,0.06707200076844957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,3584,0.03564800156487359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,4096,0.09666133589214748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,3584,0.06038755840725369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,3072,0.030565334690941706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,128,0.02976977825164795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,3584,0.08887288967768352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,3072,0.05337599913279215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,2560,0.02495199938615163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,3072,0.08068977461920844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,2048,0.039677331844965615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,1536,0.016855110724767048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,256,0.03171555532349481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,2048,0.061568889353010386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,2560,0.047671112749311656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,1024,0.012709333664841123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,2048,0.02130577796035343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,1024,0.025891555680169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,2560,0.07312977976269193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,1536,0.032960888412263654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,1536,0.052766223748524986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,768,0.01036977767944336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,1024,0.043856001562542386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,768,0.022152889106008742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,512,0.008096888661384583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,768,0.03926400012440152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,512,0.01954844428433312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,256,0.006073777874310811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,512,0.03537955549028184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,256,0.016950221525298226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,128,0.005183111048407025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,256,0.03166488806406657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,128,0.015963556038008798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,64,0.005151999907361137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3584,128,0.029467556211683486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,64,0.01605244477589925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3584,32,0.0052560000783867305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3584,32,0.01584533353646596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,16384,0.11923466788397895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,65536,0.48074844148423934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,12288,0.09362400240368313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,16384,0.27297690179612905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,10240,0.0740871098306444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,12288,0.20945955647362602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,16384,0.3199128839704725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,10240,0.17547199461195204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,65536,1.0607919692993164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,8192,0.0635404454337226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,12288,0.24867823388841417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,8192,0.139174222946167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,7168,0.05530222256978353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,10240,0.20884444978502062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,8192,0.16896533966064453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,7168,0.12061244911617702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,6144,0.04859911070929634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,6144,0.09995377726025051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,5120,0.08249155680338542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,7168,0.1482951111263699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,5120,0.041533334387673274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,6144,0.12798933188120523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,4096,0.03387199838956197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,5120,0.10798844363954331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,4096,0.06696444749832153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,3584,0.030750221676296655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,4096,0.09149689144558376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,3584,0.0604942242304484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,3072,0.026224000586403742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,3584,0.08291200134489271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,5120,512,0.037321776151657104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,3072,0.05344177616967095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,3072,0.07514311207665338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,2560,0.022663111488024395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,2560,0.04715911216206021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,2048,0.0188044442070855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,2560,0.06829688946406047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,65536,3.864723629421658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,2048,0.039023998710844256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,1536,0.0155395550860299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,2048,0.057371556758880615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,1536,0.0332915551132626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,1536,0.051685333251953125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,1024,0.011883555187119378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,1024,0.02529511186811659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,768,0.010210666391584609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,1024,0.046091556549072266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,768,0.021791110436121624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,512,0.00816533300611708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,128,0.005134222408135732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,768,0.036419557200537786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,512,0.019555555449591745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,256,0.006144000010357962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,512,0.0323022206624349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,256,0.01717688971095615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,256,0.03085600005255805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,128,0.016063110695944894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,64,0.004597333156400257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,3072,128,0.028240889310836792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,64,0.016024000114864774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,3072,32,0.004911111046870549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,3072,32,0.01568000018596649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,65536,0.40117067760891384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,16384,0.10288088851504856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,16384,0.27033599217732746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,16384,0.3138924439748128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,12288,0.07878044578764173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,12288,0.2053226629892985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,10240,0.06472533278995089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,8192,0.05435377690527174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,10240,0.17302489280700684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,12288,0.2388319969177246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,65536,2.062450620863173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,65536,1.1547457377115886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,10240,0.2040168841679891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,7168,0.04796355631616381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,8192,0.13634578386942545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,7168,0.11906222502390544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,8192,0.1653182241651747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,6144,0.048567109637790255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,7168,0.14479822582668728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,6144,0.0991520020696852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,6144,0.12503022617763943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,5120,0.03575111097759671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,5120,0.08169866932762994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,4096,0.03183733423550924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,5120,0.10633689165115356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,4096,0.06686577532026503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,3584,0.026587555805842083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,4096,0.09078222513198853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,3584,0.05987199809816149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,3072,0.024632889363500807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,3584,0.08261422316233318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,3072,0.05281244383917915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,2560,0.02052977846728431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,2560,0.04658222198486328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,3072,0.07452711131837633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,2048,0.017260443833139207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,2560,0.06729866398705377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,2048,0.039177778694364764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,2048,0.05666044685575697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,1536,0.014280888769361707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,1536,0.03263377812173631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,1024,0.011396444506115384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,1536,0.0502435564994812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,1024,0.02520622147454156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,768,0.009528000321653154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,1024,0.040884445110956825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,768,0.022202667262819078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,512,0.007735999921957652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,768,0.037081778049468994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,512,0.018941332896550495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,256,0.005817777580685086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,512,0.03181688984235128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,256,0.016713778177897137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,128,0.004962666581074397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,256,0.029719998439153034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,128,0.015782222151756287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,64,0.00452533322903845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2560,128,0.027976890405019123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,64,0.015680889288584392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2560,32,0.004763555609517627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2560,32,0.015625778171751235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,65536,0.304604450861613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,16384,0.08273066414727105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,16384,0.23384179009331596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,12288,0.0637048880259196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,12288,0.17952444818284777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,10240,0.05351555678579542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,65536,0.8956097496880425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,10240,0.15178844663831922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,8192,0.0577404432826572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,7168,0.036764442920684814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,8192,0.11983199914296468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,7168,0.10964000225067139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,6144,0.033175110816955566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,6144,0.10038222206963433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,5120,0.03646577729119195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,5120,0.0771955582830641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,16384,0.3010408878326416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,4096,0.023166222704781428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,12288,0.2311920060051812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,4096,0.057950225141313344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,10240,0.195889777607388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,3584,0.020633776982625324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,3584,0.05277511146333483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,65536,1.1267484029134114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,3072,0.018147556318177115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,8192,0.1591306659910414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,7168,0.13855111598968506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,3072,0.04606222112973531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,2560,0.016134222348531086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,6144,0.1194017728169759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,2560,0.04091555542416043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,5120,0.10230488909615411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,2048,0.013738666971524557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,2048,0.034210665358437434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,1536,0.01219911128282547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,1536,0.029422223567962646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,1024,0.00870400004916721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,1024,0.022111111217074927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,768,0.007303111255168915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,768,0.01959022217326694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,4096,0.08587733242246841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,3584,0.0787866645389133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,512,0.006120000034570694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,512,0.017456889152526855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,3072,0.0722222195731269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,256,0.004969777746333016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,2560,0.06400533517201741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,256,0.015455111861228943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,128,0.004359999878538979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,128,0.014723555909262763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,2048,0.05416799916161431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,64,0.003961777935425441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,768,0.03467466764979892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,1536,0.04773155517048306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,64,0.014565333724021912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,2048,32,0.004134222037262387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,2048,32,0.014502222339312235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,1024,0.03959111041492886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,65536,0.25527999136182994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,16384,0.06823822524812487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,16384,0.22822133700052896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,65536,0.8670390976799859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,12288,0.047431998782687716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,12288,0.17466489473978677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,10240,0.045646223757002086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,10240,0.14667999744415283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,8192,0.03209422363175286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,8192,0.1168631050321791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,128,0.02757155564096239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,7168,0.030016889174779255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,256,0.028931554820802476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,2048,512,0.030748443471060857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,7168,0.10138399733437432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,6144,0.026579555537965562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,6144,0.08470577663845485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,5120,0.02328799996111128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,5120,0.06960444317923652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,16384,0.2989484469095866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,4096,0.020404444800482858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,4096,0.05813599957360161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,12288,0.22990043958028158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,3584,0.016380444169044495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,3584,0.051579554875691734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,65536,1.1146444744533963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,3072,0.014464888307783337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,10240,0.19500088691711426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,3072,0.045774221420288086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,2560,0.012813333008024426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,2560,0.04038222299681769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,8192,0.157770660188463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,2048,0.011008888483047485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,2048,0.03337511089113023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,7168,0.13764354917738172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,1536,0.00906488878859414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,1536,0.02863911125395033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,6144,0.11894399589962429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,5120,0.10138133499357437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,1024,0.007368000017272101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,1024,0.021512000097168818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,768,0.006233777850866318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,768,0.019319999549123976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,3584,0.07914666997061835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,512,0.0053102220926019884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,2560,0.06405155526267158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,512,0.017292444904645283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,256,0.004423111263248655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,256,0.01535022258758545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,4096,0.08614310953352187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,128,0.00396799999806616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,128,0.014511111709806653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,3072,0.07180800040562947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,64,0.003689777933888965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,2048,0.053993778096305005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,64,0.01463822192615933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1536,32,0.0037644443412621817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1536,32,0.014466666513019137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,65536,0.16194844245910645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,65536,0.8719422022501627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,16384,0.05116711060206095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,16384,0.22849689589606392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,768,0.034809778134028115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,12288,0.047910223404566445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,12288,0.17170311344994438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,1536,0.04664977722697788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,10240,0.03724177678426107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,10240,0.14651466740502253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,1024,0.03780444463094076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,8192,0.03145866592725118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,8192,0.1157040066189236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,128,0.02735822233888838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,7168,0.02639644510216183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,256,0.02865333358446757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,7168,0.10207111305660671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,6144,0.023518222901556227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,6144,0.08462933037016128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,65536,1.0641618304782443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,5120,0.02072177827358246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1536,512,0.030797332525253296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,5120,0.0693733361032274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,4096,0.016946666770511203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,4096,0.05697155661053128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,16384,0.28746578428480357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,3584,0.01551822159025404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,3584,0.05134133497873942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,12288,0.21895644399854872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,3072,0.013400000002649097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,3072,0.045279999574025474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,10240,0.18530755572848848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,2560,0.011539555258221097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,2560,0.03945599993069967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,8192,0.15037066406673855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,7168,0.13176000118255615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,2048,0.009930666950013902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,2048,0.03320711188846164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,6144,0.11432888772752549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,1536,0.007714666426181793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,1536,0.02815377712249756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,1024,0.006008000009589725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,1024,0.021560000048743352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,5120,0.09744177924262153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,768,0.005442666510740916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,768,0.01939022209909227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,4096,0.08270221948623657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,512,0.004610666798220741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,3584,0.07660444577534993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,512,0.01726222203837501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,256,0.003947555604908201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,256,0.015265777707099915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,3072,0.06797689199447632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,128,0.003567999849716822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,128,0.01440000037352244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,2560,0.06061600314246284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,64,0.003343111111058129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,64,0.014370666609870063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,1024,32,0.0033822223130199644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,1024,32,0.014077333940400017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,65536,0.12396711773342556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,65536,0.8549431165059408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,2048,0.05219555563396878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,16384,0.04005777835845947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,16384,0.22677511639065215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,1536,0.0453688899676005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,12288,0.03548889027701484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,12288,0.1714773310555352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,1024,0.0363377763165368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,10240,0.03162577748298645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,10240,0.14522843890719944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,256,0.027986665566762287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,8192,0.02409066590997908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,768,0.03259822063975864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,8192,0.11675200197431777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,7168,0.02181422213713328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,7168,0.10068355666266547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,512,0.029892444610595703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,6144,0.021382222572962444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,6144,0.08336355288823445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,1024,128,0.026720888084835474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,5120,0.01719999975628323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,5120,0.06865688827302721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,65536,1.0643795861138237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,4096,0.013567999833159976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,4096,0.05611555443869697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,16384,0.28692976633707684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,3584,0.013342222405804528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,12288,0.21988177299499512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,3584,0.05127288897832235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,3072,0.011572444604502784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,3072,0.04495199852519565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,10240,0.18680711587270102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,2560,0.009550222092204625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,8192,0.15065688557094997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,2560,0.03866133424970839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,2048,0.008591111335489485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,2048,0.032855110036002264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,7168,0.13209777408176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,1536,0.006562666760550604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,1536,0.02777066661251916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,6144,0.11357155111100938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,1024,0.005403555515739653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,1024,0.021475555168257818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,5120,0.09687999884287517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,768,0.004736888739797804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,768,0.01909955508179135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,4096,0.0824106666776869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,512,0.004126222183307012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,512,0.01703555550840166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,3584,0.0745404428905911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,256,0.0035955554081334006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,256,0.015220445063379077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,3072,0.06781066788567437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,128,0.0032951111594835916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,128,0.01440533333354526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,2560,0.060031109386020236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,64,0.00309599998096625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,64,0.01426488823360867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,768,32,0.0032382222513357797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,768,32,0.014005333185195923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,65536,0.0960106650988261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,2048,0.05173333485921224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,1536,0.043037331766552396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,16384,0.03254666593339708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,65536,0.8541377915276421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,16384,0.22717600398593477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,12288,0.029194666279686823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,1024,0.03542044427659776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,12288,0.17287466261121961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,10240,0.026347556047969397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,10240,0.14517154958513048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,768,0.03210044569439358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,8192,0.02130488885773553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,8192,0.1162311103608873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,512,0.029425776667065088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,7168,0.02003466586271922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,128,0.026346666945351496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,7168,0.10048800044589573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,6144,0.015232000086042615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,768,256,0.027391110857327778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,6144,0.08358755376603867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,5120,0.015589333242840238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,5120,0.06826044453514947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,16384,0.27603199746873647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,4096,0.011581333147154914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,4096,0.05593511131074694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,65536,1.0334355036417644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,3584,0.010537777509954242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,12288,0.21270133389366996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,3584,0.05048266715473599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,3072,0.00872444444232517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,3072,0.044257776604758374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,10240,0.1813733312818739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,2560,0.007995555798212687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,2560,0.038648890124426946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,8192,0.14612621731228298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,2048,0.006807111203670502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,2048,0.0328693323665195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,7168,0.1280799971686469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,1536,0.005590222362014983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,1536,0.02665244374010298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,6144,0.1102515525288052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,1024,0.004822222308980094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,5120,0.0942408906088935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,1024,0.020981333321995203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,768,0.004317333300908406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,768,0.018768888380792405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,4096,0.079912887679206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,512,0.0038684445122877755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,512,0.016935111747847665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,3584,0.0731582244237264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,3072,0.06588533189561632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,256,0.0034017778105205963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,256,0.01480977733929952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,128,0.0031662223239739737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,128,0.014031110538376702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,2560,0.05871822436650594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,64,0.0030453333424197305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,64,0.013952000273598565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,512,32,0.0031235555393828284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,512,32,0.013636444177892474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,65536,0.06677244769202338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,65536,0.8553866810268826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,2048,0.0502782232231564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,16384,0.02276355524857839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,16384,0.22616622183057997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,1536,0.04302488764127096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,12288,0.018935999936527677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,768,0.03175466590457492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,12288,0.17176000277201334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,1024,0.03472266594568888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,10240,0.019098665979173448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,10240,0.14493332968817815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,8192,0.013918222652541267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,8192,0.11599466535780166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,512,0.02906755606333415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,7168,0.01258666647805108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,256,0.028024001253975764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,512,128,0.026211554805437725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,7168,0.10062133603625828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,6144,0.013989332649442883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,6144,0.08369244469536676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,5120,0.012435555458068848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,5120,0.06845866971545748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,65536,1.0050462086995442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,4096,0.009000889129108852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,4096,0.05617688761817085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,16384,0.26960799429151744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,3584,0.00830666638082928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,3584,0.05041155550214979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,12288,0.20731733904944527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,3072,0.007939555578761632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,3072,0.04401333464516533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,10240,0.175890670882331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,2560,0.007159111400445302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,2560,0.03853866789076064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,8192,0.1424835522969564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,2048,0.006592888798978593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,2048,0.03219288918707106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,7168,0.1252488825056288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,1536,0.005692444327804778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,1536,0.025640888346566096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,6144,0.10779288742277358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,5120,0.09195999966727363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,1024,0.004705777598751916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,1024,0.021075555019908484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,768,0.004231111043029361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,768,0.018797333041826885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,4096,0.07801244656244914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,512,0.0037324445115195382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,512,0.016689777374267578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,3584,0.07122666968239678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,256,0.0032871110985676446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,256,0.014708444476127625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,3072,0.06427022483613756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,128,0.0030213333666324615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,128,0.01389244364367591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,2048,0.048968887991375394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,2560,0.057067553202311196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,64,0.0028328889360030494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,64,0.01385155568520228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,256,32,0.0029200000895394217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,256,32,0.013569778038395775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,65536,0.05711377991570366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,1536,0.04142577780617608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,16384,0.014986667368147107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,65536,0.8565004666646322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,1024,0.034104890293545194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,768,0.031339555978775024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,16384,0.22712445259094238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,12288,0.01126933346192042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,10240,0.01070755554570092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,12288,0.17255022790696886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,10240,0.14470399750603571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,8192,0.008569777839713627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,8192,0.1165448824564616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,512,0.028730667299694482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,7168,0.012058667010731168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,7168,0.10103288624021743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,256,0.02686400049262577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,6144,0.008614222208658854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,6144,0.0831200016869439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,256,128,0.026260443859630164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,5120,0.008945778012275696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,5120,0.06755022207895915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,65536,1.001198238796658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,16384,0.2692008813222249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,4096,0.008178667061858708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,10240,0.17493422826131186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,4096,0.05548266569773356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,12288,0.2075351079305013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,3584,0.008466666771305932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,3584,0.05057422320048014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,3072,0.007854222423500484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,3072,0.0439573327700297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,2560,0.0063742221229606206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,2560,0.0380675560898251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,8192,0.14266310797797307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,2048,0.005522666705979242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,2048,0.031883554326163396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,7168,0.12457777394188775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,1536,0.004859555512666702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,1536,0.02553155521551768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,6144,0.1070186694463094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,1024,0.004171555654870139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,1024,0.020845333735148113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,5120,0.0912968913714091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,768,0.0038933331767717996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,768,0.018505778577592637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,4096,0.07750666803783841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,512,0.0035217776894569397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,3072,0.06379289097256131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,512,0.016543111867374845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,256,0.0031511110977994073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,3584,0.07156622409820557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,256,0.014576888746685453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,128,0.00291377781993813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,2560,0.05646666553285387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,128,0.013799111048380533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,64,0.0027884443600972495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,64,0.01370666672786077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,128,32,0.0028151110228565005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,128,32,0.013366222381591797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,65536,0.05562488900290596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,2048,0.048411554760403104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,16384,0.011920889218648275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,65536,0.8548773129781088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,16384,0.22625332408481172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,12288,0.009842666486899057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,10240,0.008404444489214156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,12288,0.17186666859520805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,8192,0.007541333635648091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,10240,0.1445137792163425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,7168,0.007968000239796108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,8192,0.11650399367014568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,6144,0.0086666668454806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,7168,0.10035288996166652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,6144,0.08316089047325982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,5120,0.008047999607192146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,4096,0.007375111182530721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,5120,0.06827288866043091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,1536,0.04012889001104567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,3584,0.0070266665683852295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,4096,0.05569510989718967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,3584,0.04974044362703959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,3072,0.006684444430801604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,3072,0.04402133491304186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,2560,0.006186666587988536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,2560,0.037316444847318865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,2048,0.005449777675999536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,1536,0.004741333425045013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,2048,0.031193776263131037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,1536,0.025391111771265667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,1024,0.004056000047259861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,1024,0.020954666866196528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,768,0.0037466668420367767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,768,0.018657777044508193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,512,0.003400000019205941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,512,0.01662133302953508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,256,0.003060444361633725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,256,0.014656888114081489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,128,0.002826666666401757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,128,0.013889777991506787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,64,0.002891555635465516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,64,0.013745778136783175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,64,32,0.002887111157178879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,64,32,0.0135057775510682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,65536,0.053416000472174756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,16384,0.010616000327799054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,16384,0.22606044345431855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,12288,0.009152889251708984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,65536,0.855439133114285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,12288,0.17190044456058076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,10240,0.008356444537639618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,8192,0.007382222347789341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,10240,0.14453066719902888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,7168,0.00721066693464915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,8192,0.11619022157457139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,7168,0.10014933347702026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,6144,0.006493333313200209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,5120,0.006798222247097228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,6144,0.08284355534447564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,1024,0.03376177615589566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,4096,0.006456888798210356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,5120,0.06777422295676337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,3584,0.006822222222884496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,4096,0.05510044429037306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,3072,0.006435555716355641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,3584,0.05020799901750353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,3072,0.043495999442206494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,2560,0.006050666587220297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,2560,0.03772177630000644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,2048,0.0053431110249625305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,2048,0.030047999487982854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,1536,0.004695999953481886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,1536,0.025412445267041523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,1024,0.003979555434650845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,1024,0.02087733315096961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,768,0.003640000190999773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,768,0.01853155593077342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,512,0.0033484444850020935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,512,0.016560888952679105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,256,0.0029893333299292457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,256,0.014736000034544202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,128,0.0027760000278552375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,128,0.013863110707865821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,64,0.01370044466522005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,64,0.0026266665922270883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1024,32,32,0.002790222151411904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1024,32,32,0.013420444395807056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,16384,1.3609547085232203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,12288,1.4893688625759547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,16384,2.2169698079427085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,12288,1.03438843621148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,768,0.03125955661137899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,10240,0.8678604231940376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,10240,1.850687026977539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,8192,0.9728533426920573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,8192,0.9014577865600586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,16384,1.3570017284817164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,512,0.02884355518552992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,12288,1.004651599460178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,7168,1.3049892849392362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,256,0.02721600068940057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,7168,0.5988924238416884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,6144,0.753975126478407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,10240,0.8321573469373914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,6144,0.7591635386149088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,5120,0.6588995721605089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,5120,0.4206382168663873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,8192,0.6767217848036023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,4096,0.3436275588141547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,4096,0.8053395483228895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,7168,0.5877075725131565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,3584,0.4502879778544108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,3584,0.2989857726626926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,6144,0.5112346543206109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,3072,0.3682106600867377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,5120,0.4299395614200168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,3072,0.25889955626593697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,2560,0.31152621905008954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,4096,0.3524613380432129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,2560,0.21285600132412383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,2048,0.26464888784620494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,3072,0.27493688795301646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,2048,0.17518488566080728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,1536,0.19463645087348092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,1536,0.1355866723590427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,2560,0.23674488067626953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,1024,0.13974489106072321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,1024,0.0993564460012648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,768,0.11782666047414143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,3584,0.7797733412848579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,1536,0.16251556078592935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,2048,0.197272883521186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,1024,0.12188088893890381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,768,0.08497777912351821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,512,0.08475555313958062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,512,0.07160266902711657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,768,0.10548266437318589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,256,0.05804889069663154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,512,0.08619466755125259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,128,0.05660533242755466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,256,0.060921774970160596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,64,0.045237332582473755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,128,0.07400711377461751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,64,0.05902489026387533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,128,0.050027555889553495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,65536,256,0.07874044444825914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,65536,32,0.04562044474813673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,65536,32,0.060495111677381724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,16384,0.47773064507378477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,16384,0.42831910981072324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,12288,0.35370577706231016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,12288,0.32759467760721844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,65536,1.716915554470486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,65536,1.9479351043701172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,10240,0.30125244458516437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,16384,0.46602312723795575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,10240,0.27542222870720756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,8192,0.25358578893873424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,8192,0.21167733934190539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,12288,0.35935378074645996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,7168,0.1998924414316813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,10240,0.29680710368686253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,7168,0.18793422645992705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,6144,0.17950933509402803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,6144,0.15976533624860975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,7168,0.21504088242848715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,8192,0.2422382301754422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,5120,0.15985243850284153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,5120,0.1333120001686944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,6144,0.1852960056728787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,4096,0.12674044238196477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,5120,0.16236266824934217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,4096,0.10936178101433648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,3584,0.10953511132134332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,3584,0.09797155857086182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,4096,0.13716977172427708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,3072,0.09527022308773464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,3072,0.08561066786448161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,3584,0.12381245030297174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,2560,0.08702133099238078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,3072,0.10935999949773152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,2560,0.07291999790403578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,2048,0.06833777825037639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,2048,0.06143555376264784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,2560,0.09640800290637547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,1536,0.055383112695482045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,2048,0.0838444431622823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,1536,0.05158844590187073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,1024,0.040236443281173706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,1536,0.0718640022807651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,1024,0.039480000734329224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,768,0.0336079994837443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,1024,0.05743822124269274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,768,0.052575111389160156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,512,0.030375109778510198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,768,0.03529689047071669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,512,0.02588622272014618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,256,0.01796444422668881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,512,0.045679999722374805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,256,0.024864888853496973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,128,0.014215999179416232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,256,0.041067557202445135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,64,0.022750222020679053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,128,0.023543111152119104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,64,0.01257599973016315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,128,0.03848177856869168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,16384,32,0.013040888640615674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,16384,32,0.022985777921146814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,16384,0.3306826750437419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,16384,0.3748782210879856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,12288,0.25025423367818195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,12288,0.2836995654635959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,65536,1.491793738471137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,65536,1.4869662390814886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,10240,0.22000000211927626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,10240,0.24944178263346353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,8192,0.1730355554156833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,8192,0.18349866072336832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,7168,0.16230577892727321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,7168,0.16003110673692492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,16384,65536,7.564879523383246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,6144,0.6692595481872559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,6144,0.4629075792100694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1024,128,128,0.025768889321221247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,5120,0.11278488900926377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,5120,0.11585777335696751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,12288,0.3053804503546821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,4096,0.09268889162275527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,4096,0.09534311294555664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,16384,0.4001164436340332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,3584,0.08511377705468072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,10240,0.25282133950127494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,3584,0.08414577775531346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,8192,0.2068613370259603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,65536,1.5650044547186956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,3072,0.07420800129572551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,3072,0.07713599999745686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,2560,0.06423822376463148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,2560,0.06463999880684747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,7168,0.18300444549984404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,2048,0.05117866728040907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,2048,0.054162667857276074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,1536,0.04013599952061971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,1536,0.04608266552289327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,6144,0.19568355878194174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,1024,0.0418835547235277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,1024,0.044907556639777295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,5120,0.13802844948238796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,768,0.024105777343114216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,768,0.03073244293530782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,3584,0.10549600256813897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,512,0.018326222896575928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,512,0.026709333062171936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,2560,0.08451733324262832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,256,0.013160000244776407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,256,0.02311733365058899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,4096,0.1166799995634291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,128,0.010093332992659675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,128,0.021028444170951843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,1536,0.06452977657318115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,3072,0.09414844380484687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,64,0.008392000363932716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,64,0.0211173329088423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,12288,32,0.008656889200210571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,12288,32,0.020451555649439495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,2048,0.07302133242289226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,65536,1.2538044187757704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,65536,1.291094250149197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,16384,0.2940853436787923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,16384,0.3231724368201362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,12288,0.23510755432976616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,12288,0.2480248875088162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,1024,0.05098933312628004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,10240,0.27452267540825737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,10240,0.2333830992380778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,768,0.04749777913093567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,8192,0.15912266572316489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,8192,0.1577955616845025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,128,0.03576711151334975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,7168,0.1364311112297906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,256,0.038931555218166776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,7168,0.1385706663131714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,6144,0.11605778005388047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,6144,0.12031466431087917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,12288,512,0.04138133260938857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,5120,0.10255377822452122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,5120,0.09955200221803452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,16384,0.383802678849962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,12288,0.28643645180596244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,4096,0.08298489120271471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,4096,0.08371200164159139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,3584,0.07595733139250013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,3584,0.07455822494294909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,65536,1.4211644066704645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,3072,0.06651822063657972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,3072,0.068031112353007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,10240,0.262718227174547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,2560,0.0663164456685384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,2560,0.05889955494138929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,8192,0.1945128838221232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,2048,0.045421332120895386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,2048,0.04875466558668348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,7168,0.17300266689724395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,1536,0.036044445302751325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,6144,0.15143822299109563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,1536,0.04182489050759209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,1024,0.025742222865422566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,1024,0.032016889916525945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,5120,0.13106844160291883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,768,0.021200888686709937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,768,0.027895109521018133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,3584,0.10082933637830947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,512,0.016568000117937725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,512,0.02482755482196808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,4096,0.11089866691165501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,256,0.011615110768212212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,256,0.020823111136754353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,3072,0.09052799807654487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,128,0.010800888968838586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,128,0.018791110979186166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,2560,0.07910222477383085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,64,0.009708444277445475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,64,0.019000889526473153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,10240,32,0.010088889135254754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,10240,32,0.018763555420769584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,65536,1.217930687798394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,65536,1.312604480319553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,2048,0.06958488623301189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,16384,0.23805422253078887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,16384,0.2795582347446018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,1536,0.06077510780758328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,12288,0.17028710577223036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,768,0.04478489028082954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,12288,0.21505067083570692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,10240,0.1429493294821845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,512,0.0398888885974884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,10240,0.1815048853556315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,8192,0.11557866467369927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,128,0.03321511215633816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,256,0.0370417767100864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,8192,0.13671199480692545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,7168,0.10281955533557469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,7168,0.11957244078318278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,6144,0.09267289108700222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,6144,0.10509422090318467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,10240,1024,0.04788088798522949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,5120,0.08024888568454318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,5120,0.08644977543089126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,65536,1.988726298014323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,4096,0.09265333414077759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,4096,0.07417244381374784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,16384,0.3297671212090386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,3584,0.05942044655481974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,3584,0.06388888756434123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,10240,0.21247555149926078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,3072,0.0518506666024526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,3072,0.0586986674202813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,12288,0.251855108473036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,2560,0.0428711109691196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,2560,0.048912889427608915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,6144,0.133206221792433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,7168,0.15327466858757868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,2048,0.035156445370780096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,2048,0.04187644521395365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,1536,0.0270471109284295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,8192,0.1715759966108534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,1536,0.03716444306903415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,5120,0.1156346665488349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,1024,0.019887111253208585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,1024,0.028563555743959215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,768,0.016466667254765827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,768,0.025219556358125474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,3584,0.09014311101701523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,512,0.013312888642152151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,512,0.022610666023360357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,4096,0.09868622488445705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,768,0.04235288831922743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,256,0.011767999993430244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,256,0.019504000743230183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,128,0.007420444654093848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,128,0.017311111092567444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,2560,0.07105333275265165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,64,0.006424888968467712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,64,0.017699556218253244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,8192,32,0.006683555742104848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,8192,32,0.017398221625222098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,65536,0.8239902390374078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,1536,0.05677600039376152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,65536,1.0881004333496094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,3072,0.0813466641638014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,16384,0.20686754915449354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,16384,0.2751244439019097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,12288,0.15618933571709528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,2048,0.06242933538224962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,12288,0.2105191151301066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,10240,0.12916799386342367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,10240,0.17768177721235487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,1024,0.0447599987188975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,8192,0.10288800133599176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,8192,0.13407288657294378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,128,0.031457778480317854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,7168,0.10017777813805474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,7168,0.11724978023105198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,256,0.035544001393847995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,6144,0.0890773336092631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,6144,0.10471200280719334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,8192,512,0.037780443827311196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,5120,0.07148089011510213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,5120,0.08553244670232137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,65536,1.2176062266031902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,16384,0.32462578349643284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,4096,0.0612915555636088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,12288,0.24844000074598524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,4096,0.07000444332758586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,10240,0.2085404396057129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,3584,0.0502017769548628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,3584,0.0633173320028517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,3072,0.04265866676966349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,3072,0.057518223921457924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,2560,0.03821422325240241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,2560,0.047761778036753334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,8192,0.16894043816460502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,2048,0.03163733416133457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,2048,0.04103822178310818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,7168,0.14967111746470133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,1536,0.024863110648261175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,1536,0.03606755534807841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,6144,0.13174488809373644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,1024,0.019168888529141743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,1024,0.0289066665702396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,5120,0.1128773291905721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,768,0.01521511044767168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,768,0.025095999240875244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,3584,0.08746755785412258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,512,0.011438222395049201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,512,0.02179733415444692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,2560,0.0700026684337192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,256,0.008973333570692275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,256,0.018403554956118267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,4096,0.09666044182247585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,128,0.007650666766696506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,128,0.017343999611006845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,3072,0.07978933387332492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,64,0.007335999773608313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,64,0.01775911119249132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,7168,32,0.007716444631417592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,7168,32,0.017292444904645283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,65536,0.7453218036227756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,65536,1.0749137666490343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,2048,0.061477336618635386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,16384,0.168649779425727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,16384,0.27080000771416557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,1536,0.05598577525880602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,12288,0.1361208889219496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,12288,0.20684888627794054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,768,0.042035556501812406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,10240,0.11435288853115505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,10240,0.17405777507358125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,1024,0.04402044415473938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,8192,0.09781955348120795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,8192,0.13083467218610975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,128,0.031903998719321355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,7168,0.08182044161690606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,256,0.0332622230052948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,7168,0.11510311232672797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,6144,0.07006666395399305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,6144,0.09994844595591228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,7168,512,0.03739555676778158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,5120,0.0603333314259847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,5120,0.08320622311698066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,65536,1.246188481648763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,16384,0.3326408863067627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,4096,0.04941866795221964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,4096,0.06872800323698255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,3584,0.04129688938458761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,3584,0.06139555242326525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,12288,0.25283556514316136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,3072,0.037949333588282265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,3072,0.05718666977352566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,10240,0.21213600370619032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,2560,0.03216444452603658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,2560,0.04741866721047295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,8192,0.17138844066196016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,2048,0.027868443065219458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,2048,0.040116445885764226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,7168,0.150144894917806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,1536,0.021568000316619873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,1536,0.034308443466822304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,6144,0.13050933678944907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,1024,0.016232000456915963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,1024,0.027619555592536926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,5120,0.11379644605848525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,768,0.013632888595263163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,768,0.02462844385041131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,4096,0.0967902210023668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,512,0.010996444357766045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,512,0.020997333857748244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,3584,0.08852977885140313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,256,0.007800888684060838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,256,0.018079999420377944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,3072,0.07918666468726264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,128,0.006518222391605377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,128,0.017127111554145813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,2560,0.0698231127527025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,64,0.005619555711746216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,64,0.017255110873116385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,6144,32,0.005985777825117111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,6144,32,0.017117333081033494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,2048,0.06135644515355428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,65536,0.5923982196384007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,1536,0.05385422375467089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,65536,0.9068124559190539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,16384,0.1540044413672553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,1024,0.044751998451020986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,16384,0.22946667671203613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,12288,0.11593955092959934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,12288,0.1773244407441881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,10240,0.13618132803175184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,10240,0.17144355509016249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,768,0.042174223396513194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,8192,0.08070755667156644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,8192,0.10930399762259589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,256,0.03355111016167535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,7168,0.070342222849528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,7168,0.09686222341325547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,512,0.0366284449895223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,6144,0.062084443039364286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,6144,0.08443555566999648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,6144,128,0.031195554468366835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,5120,0.052404443422953285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,5120,0.06974311007393731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,16384,0.2838497691684299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,65536,1.0485973358154297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,4096,0.042452444632848106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,12288,0.21782844596438936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,4096,0.05876355701022678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,3584,0.0381315549214681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,3584,0.05381422241528829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,3072,0.033915556139416166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,10240,0.19151555167304146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,3072,0.04963733421431648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,2560,0.029151999288135107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,2560,0.041434665520985924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,8192,0.14758222632937962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,2048,0.024069334069887798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,2048,0.03539022141032749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,6144,0.11569689379798041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,1536,0.019259555472267997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,7168,0.13276178307003444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,1536,0.03055022160212199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,1024,0.014653333359294467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,1024,0.024518221616744995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,5120,0.10074577728907268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,768,0.012501333322789935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,768,0.021886222892337378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,3584,0.07874311341179742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,512,0.009937778115272522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,512,0.018862222631772358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,2560,0.06348799996905856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,256,0.00721066693464915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,256,0.016770665844281513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,4096,0.0856471127933926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,128,0.006067555397748947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,128,0.015779554843902588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,3072,0.07247110870149401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,1536,0.04979111088646782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,64,0.0052684446175893145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,5120,32,0.005398222141795688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,64,0.016063110695944894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,5120,32,0.015601777368121676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,65536,0.4487039777967665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,2048,0.05526933405134413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,65536,0.8928889168633355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,16384,0.11933333343929714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,16384,0.23098932372199166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,12288,0.09481333361731635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,12288,0.1729733281665378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,1024,0.04052799940109253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,10240,0.07384355862935384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,10240,0.14456710550520155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,768,0.03898222247759501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,8192,0.06355822086334229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,8192,0.10990666680865818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,256,0.03218933277659946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,7168,0.055799113379584424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,512,0.03352800011634827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,7168,0.09519288937250774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,6144,0.05046844482421875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,6144,0.08286844359503852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,5120,128,0.03015466531117757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,5120,0.042583111259672374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,5120,0.06900088654624091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,16384,0.2621884346008301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,4096,0.03449066811137729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,4096,0.058113780286577016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,12288,0.20300711525811088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,3584,0.03214933474858602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,3584,0.053815109862221606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,65536,0.9777342478434244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,3072,0.026802667313151892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,3072,0.04820977648099264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,10240,0.1687999963760376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,2560,0.022430222895410325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,2560,0.040639999839994646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,8192,0.13592888249291316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,2048,0.019110222657521565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,2048,0.03540622194608053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,7168,0.12177599800957574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,6144,0.10637955533133613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,1536,0.015410666664441427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,1536,0.03052355514632331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,1024,0.011744000017642975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,1024,0.02409155501259698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,5120,0.09254222446017796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,768,0.009848000274764167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,768,0.02121777832508087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,3584,0.07431288560231526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,512,0.007885333564546373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,512,0.018627555833922494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,4096,0.0801911089155409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,256,0.006079111248254776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,256,0.016813332835833233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,3072,0.06766488817003039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,128,0.005210666606823604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,128,0.015863110621770222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,2560,0.05976088841756185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,64,0.004663111021121343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,64,0.015993777248594496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,4096,32,0.00490133340160052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,4096,32,0.015604444675975375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,65536,0.38517154587639707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,2048,0.05369244350327385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,65536,0.8735635545518664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,16384,0.10205777486165364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,16384,0.222725338406033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,768,0.0373582210805681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,12288,0.08152711391448975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,12288,0.171052442656623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,1536,0.04724444283379448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,10240,0.0690808892250061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,10240,0.14303110705481634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,1024,0.03888266616397434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,8192,0.055345777008268565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,8192,0.108115553855896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,256,0.03076710965898302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,7168,0.04942933387226529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,128,0.02952444553375244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,7168,0.09459288914998372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,6144,0.04289244280921089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,6144,0.08221777942445543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,4096,512,0.03287200133005778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,5120,0.037864890363481306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,5120,0.06904355684916179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,16384,0.259190215004815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,4096,0.03038933210902744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,4096,0.057651552889082164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,65536,0.9592293633355035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,3584,0.027965333726671007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,12288,0.19863643911149767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,3584,0.05236533284187317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,3072,0.024167110522588093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,3072,0.04823377728462219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,10240,0.16722488403320312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,2560,0.020767110917303298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,2560,0.040122667948404946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,8192,0.13454043865203857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,2048,0.01756800048881107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,2048,0.0351493325498369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,6144,0.10533599721060859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,1536,0.014495111174053617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,1536,0.03020533257060581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,7168,0.12081866794162327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,1024,0.011536888778209686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,1024,0.023775999744733173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,5120,0.09246933460235596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,768,0.009835555321640438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,768,0.02033511135313246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,3584,0.07305422094133165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,512,0.007919111185603673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,512,0.018193778064515855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,2560,0.05801688962512546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,256,0.006342222293217977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,4096,0.07928178045484754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,256,0.01644888851377699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,128,0.005482666608360078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,128,0.015646222564909194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,3072,0.0671022203233507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,64,0.005165333135260476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,64,0.0159262220064799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3584,32,0.005444444302055571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3584,32,0.015650666422314115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,65536,0.3487991227044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,65536,0.868959108988444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,2048,0.053054223457972206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,16384,0.0927964448928833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,16384,0.22211200661129424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,768,0.037095109621683754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,1536,0.0462773342927297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,12288,0.07387377818425496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,12288,0.17053422662946913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,10240,0.06197600232230293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,10240,0.14223733213212755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,1024,0.038500444756613836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,8192,0.04888355400827196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,8192,0.107588443491194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,128,0.02943377693494161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,7168,0.043024000194337636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,7168,0.09431733025444879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,256,0.030476444297366675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,6144,0.038132445679770574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,6144,0.08192000124189588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3584,512,0.03154844376775954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,5120,0.034612443712022566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,5120,0.06895822286605835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,65536,0.9735298156738281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,4096,0.026693332526418898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,4096,0.057822220855289035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,16384,0.2606168852912055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,3584,0.023544000254737005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,3584,0.05217155483033922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,10240,0.1672346724404229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,12288,0.2006782160864936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,3072,0.021914665897687275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,3072,0.047212445073657565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,2560,0.018374222848150466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,2560,0.03955111238691542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,8192,0.13556978437635633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,2048,0.015288000305493673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,2048,0.034648888640933566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,7168,0.11899644798702663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,1536,0.012539555629094442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,6144,0.10470310846964519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,1536,0.02987733483314514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,1024,0.009932444327407414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,1024,0.022130666507614985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,5120,0.09096266825993855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,768,0.008344888687133789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,768,0.020041777027977836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,4096,0.08002755377027723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,512,0.006782222125265334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,512,0.018583999739752877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,3584,0.0735679997338189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,256,0.0053297777970631914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,256,0.016366221838527255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,3072,0.06690044535530938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,128,0.004601777841647466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,2560,0.05835821893480089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,128,0.015528000063366361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,64,0.0042444442709287005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,64,0.01566488875283135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,3072,32,0.004400889078776042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,3072,32,0.015407999356587728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,65536,0.3111831082238091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,65536,0.7133164405822754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,2048,0.05289599961704678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,16384,0.07806044154696994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,16384,0.18458578321668836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,1536,0.04694844285647074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,12288,0.06025155385335287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,12288,0.14088000191582575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,1024,0.03899733225504557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,768,0.03560355636808608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,10240,0.051295999023649425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,10240,0.11737955941094293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,8192,0.041972445117102734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,8192,0.08936444256040786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,256,0.030842665168974135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,7168,0.03721155391799079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,512,0.03209777673085531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,7168,0.07879110839631823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,6144,0.032666666640175715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,6144,0.06831555234061347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,3072,128,0.02936977810329861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,5120,0.028880000114440918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,5120,0.05793866846296522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,65536,0.8946684731377496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,4096,0.023390221926901076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,4096,0.048519111341900296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,16384,0.24279022216796875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,3584,0.020996444755130347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,3584,0.04462488823466831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,12288,0.18672088781992593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,3072,0.0184515549076928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,3072,0.04075555668936835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,10240,0.15545866224500868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,2560,0.015983111328548856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,2560,0.034126222133636475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,8192,0.1250577767690023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,2048,0.01365066650840971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,2048,0.03016177813212077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,7168,0.11208800474802653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,1536,0.01148977792925305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,1536,0.027190221680535212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,6144,0.09810222519768609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,1024,0.008986666798591614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,1024,0.02040177749262916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,5120,0.08628888924916585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,768,0.007722666694058313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,768,0.018346667289733887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,4096,0.07461688915888469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,512,0.006160000132189856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,512,0.016764443781640794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,3584,0.06931022140714857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,256,0.005029333134492238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,256,0.015293333265516492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,3072,0.06276444594065349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,128,0.004369777937730153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,128,0.014456000592973499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,2560,0.05473688907093472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,64,0.0041555555330382455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,64,0.014540443817774454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2560,32,0.004238222208287981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2560,32,0.01423733267519209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,2048,0.04879200127389696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,65536,0.23335109816657174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,1536,0.04465244544876946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,16384,0.09433155589633518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,16384,0.18410400549570718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,1024,0.035869333479139544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,65536,0.6984720230102539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,12288,0.0733955568737454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,12288,0.13976444138420954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,10240,0.06020800272623698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,10240,0.11642577913072373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,768,0.033385776811175875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,8192,0.045554667711257935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,8192,0.08880800008773804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,256,0.029654221402274236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,7168,0.04208533300293816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,512,0.030504888958401147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,7168,0.07704355319341023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,6144,0.036694222026401095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,6144,0.06708711385726929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2560,128,0.028331554598278467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,5120,0.03094666533999973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,5120,0.057188444667392306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,12288,0.18781867292192247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,16384,0.24299287796020508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,4096,0.021094222863515217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,4096,0.04840444525082906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,3584,0.018079110317760043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,65536,0.9011902279324002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,3584,0.04440800017780728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,3072,0.016656888855828177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,3072,0.04007555709944831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,10240,0.15616356001959905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,2560,0.014549333188268872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,2560,0.03398222062322829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,8192,0.12582133875952825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,2048,0.012469333079126147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,2048,0.029792000850041706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,7168,0.11141422059800889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,1536,0.009970666633711921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,1536,0.02644444505373637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,6144,0.09801689121458267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,1024,0.00758755538198683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,1024,0.01998488936159346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,5120,0.08636089166005452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,768,0.006302222195598815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,768,0.018218666315078735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,3584,0.06927555799484253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,512,0.005331555588377847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,512,0.016710221767425537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,2560,0.05520088805092705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,256,0.004440000073777305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,256,0.015073777900801765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,4096,0.07490489217970106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,128,0.003920000046491623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,128,0.01440711153878106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,3072,0.06254844533072577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,64,0.003584889074166616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,64,0.0144177774588267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,2048,32,0.003809777812825309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,2048,32,0.014180445008807711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,65536,0.1904026667277018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,65536,0.6993430985344781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,2048,0.04924622178077698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,16384,0.05764800310134888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,1536,0.04471733172734579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,16384,0.18369689252641466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,12288,0.04176977939075894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,12288,0.13895732826656765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,768,0.033152888218561806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,10240,0.03513866662979126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,10240,0.11448089281717937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,1024,0.036462220880720354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,8192,0.027783112393485174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,8192,0.08852977885140313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,128,0.028295109669367473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,256,0.029254221253924902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,7168,0.024950222836600408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,7168,0.07678311400943331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,6144,0.021405332618289526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,6144,0.06622222397062513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,2048,512,0.030280888080596924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,5120,0.018802666001849704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,5120,0.05735466877619425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,65536,0.8817848629421658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,4096,0.015767999821239047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,4096,0.04828177889188131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,16384,0.23946756786770293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,3584,0.01425866617096795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,3584,0.04313155677583483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,12288,0.18490133020612928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,3072,0.012903999951150684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,3072,0.03908088803291321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,10240,0.1539342270957099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,2560,0.011335999601417117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,2560,0.033864001433054604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,8192,0.1240088939666748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,2048,0.012523555921183692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,2048,0.029806223180558946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,7168,0.11064710881974961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,1536,0.01001422190003925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,1536,0.025974222355418738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,6144,0.09691111246744792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,1024,0.0075697774688402815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,1024,0.020091555184788175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,5120,0.08558666706085205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,768,0.006423111177153057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,768,0.018228444788191054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,4096,0.07421600156360202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,512,0.005170666509204441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,3584,0.06805421908696492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,512,0.016933333542611863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,256,0.0041031113101376426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,256,0.015288889408111572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,3072,0.06167111131880018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,128,0.003644444462325838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,2560,0.05489066574308607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,128,0.014304000470373364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,64,0.0034622223012977173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,64,0.014304889572991265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1536,32,0.0035955554081334006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1536,32,0.014007111390431723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,65536,0.12345155080159505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,65536,0.6874088711208768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,2048,0.048835555712381996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,16384,0.04020977682537503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,16384,0.18254755602942574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,1536,0.044363554981019765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,12288,0.052431109878751964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,12288,0.13830933305952284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,1024,0.03513955407672458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,10240,0.04264800084961785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,10240,0.11480355262756348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,768,0.0336079994837443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,8192,0.034727109803093806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,8192,0.08835911088519627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,256,0.029300444655948218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,7168,0.03139999839994643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,7168,0.07688533597522311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,512,0.0299911101659139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1536,128,0.028127110666698877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,6144,0.026165333059098985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,6144,0.0669777790705363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,5120,0.016935111747847665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,5120,0.057189332114325635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,65536,0.8414142396714953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,4096,0.013663999736309052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,4096,0.04713777701059977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,16384,0.22919644249810112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,3584,0.011750222080283694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,3584,0.04254133171505398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,12288,0.17566843827565512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,3072,0.01167377746767468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,3072,0.038429333104027644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,10240,0.14672799905141196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,2560,0.009481777747472128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,8192,0.1186293363571167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,2560,0.03303822212749057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,2048,0.00869511150651508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,2048,0.02864444586965773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,6144,0.09355822536680435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,1536,0.006634666687912411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,7168,0.10506844520568848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,1536,0.024283554818895128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,5120,0.08192711406283908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,1024,0.0054560001525614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,1024,0.019680889116393197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,768,0.004822222308980094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,768,0.018041777941915724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,4096,0.07015466690063477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,512,0.004181333300140169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,512,0.01640977793269687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,3584,0.06457955307430692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,256,0.003681777666012446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,256,0.01497866710027059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,768,0.03174044357405768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,128,0.0033688888781600525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,128,0.014128888646761576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,2560,0.0518506666024526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,64,0.0031795555518733132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,64,0.01418133411142561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,3072,0.05810666746563382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,1024,32,0.0033128888656695685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,1024,32,0.013929777675204806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,65536,0.10422044330173069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,1536,0.040229333771599665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,16384,0.03363200028737386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,2048,0.04618577824698555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,65536,0.6821244557698568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,16384,0.1796462270948622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,12288,0.027621333797772724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,12288,0.13608800040351018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,10240,0.023316444622145757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,1024,0.03295644455485874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,10240,0.11393777529398601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,8192,0.018770666586028207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,8192,0.08873244126637776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,256,0.028632889191309612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,7168,0.017055999901559617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,512,0.028892444239722356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,7168,0.0763840013080173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,6144,0.014952888091405233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,6144,0.06586488750245836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,1024,128,0.027608889672491286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,5120,0.013367999758985309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,5120,0.05626489056481255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,16384,0.22192266252305773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,12288,0.17098044024573433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,65536,0.8205973307291666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,4096,0.010904889139864178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,4096,0.04716711242993673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,10240,0.14322933885786268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,3584,0.010041777458455827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,3584,0.042633778519100614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,8192,0.1153528822792901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,3072,0.00903555585278405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,3072,0.03836355606714884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,2560,0.01147911118136512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,2560,0.03338844577471415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,2048,0.009595555563767752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,2048,0.028943111499150593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,7168,0.10286666949590047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,1536,0.00793422261873881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,1536,0.02369333306948344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,6144,0.0905911127726237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,1024,0.006093333164850871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,1024,0.01942488882276747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,5120,0.08015288909276326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,768,0.005203555441564984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,768,0.01778044468826718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,3584,0.063573333952162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,4096,0.0690880020459493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,512,0.004415110995372136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,512,0.016144000821643405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,256,0.0037004442678557504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,3072,0.05749866697523329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,256,0.014628445108731588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,128,0.003350222276316749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,128,0.01391377713945177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,2560,0.051128000020980835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,64,0.003154666680428717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,64,0.013889777991506787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,768,32,0.003144888828198115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,768,32,0.01368799971209632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,65536,0.07687289184994169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,2048,0.045165333482954234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,65536,0.6821902063157824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,16384,0.0274008893304401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,16384,0.18141867054833305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,1536,0.04066844450102912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,12288,0.020958221620983545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,12288,0.13584444257948133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,1024,0.032749334971110024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,10240,0.0199937770764033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,10240,0.11416888236999512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,768,0.03173155585924784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,8192,0.016963554753197562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,8192,0.08703466918733384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,256,0.028556442923016016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,512,0.02875022093454997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,7168,0.01497866710027059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,7168,0.07641777727339003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,6144,0.013320888910028668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,768,128,0.027387556102540758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,6144,0.06567466921276517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,5120,0.01180622229973475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,5120,0.05609155363506741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,65536,0.8108631239997016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,4096,0.01015644437736935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,16384,0.21852621767255995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,4096,0.046606222788492836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,3584,0.009205333060688442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,3584,0.041907555527157254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,12288,0.16812445057762992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,3072,0.008553778131802877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,3072,0.038223998414145574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,10240,0.14046043819851345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,2560,0.007229333122571309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,2560,0.031515555249320135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,8192,0.11373066902160645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,2048,0.006138666636413998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,2048,0.028023110495673284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,7168,0.10066755612691243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,1536,0.005474666754404704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,1536,0.023803555303149756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,6144,0.08874222305085923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,1024,0.004569777597983678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,1024,0.019115555617544387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,5120,0.078311112191942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,768,0.0041706665522522396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,768,0.017627555463049147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,4096,0.06744088729222615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,512,0.0037457777394188773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,512,0.01606044504377577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,3584,0.06178222099939982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,256,0.0033440000067154565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,256,0.014563555518786112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,3072,0.056186669402652316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,128,0.003080000066094928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,128,0.013838222457302941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,2560,0.04980088935958015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,64,0.0029075555503368378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,64,0.013824889229403602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,512,32,0.003063999944263034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,512,32,0.013615111509958902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,65536,0.05602400170432197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,65536,0.6821413569980197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,2048,0.04451377855406868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,16384,0.017997332745128207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,16384,0.18129777908325195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,1536,0.03889244463708665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,12288,0.01588000026014116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,12288,0.13761156135135227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,1024,0.032433778047561646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,768,0.03139288889037238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,10240,0.014430221584108142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,10240,0.11413421895768906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,8192,0.0125591109196345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,512,0.02866311205757989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,8192,0.08784355719884236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,7168,0.01163644426398807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,256,0.027944889333513048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,7168,0.07672622468736437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,6144,0.012060444388124677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,6144,0.0659066637357076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,512,128,0.027425777581002977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,5120,0.010224888722101847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,5120,0.055360890097088285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,65536,0.7775475713941785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,4096,0.008769777913888296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,16384,0.2119928863313463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,4096,0.0466284453868866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,3584,0.008013333711359236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,3584,0.04173955652448866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,12288,0.16266222794850668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,3072,0.007152000235186682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,10240,0.1358613305621677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,3072,0.03757155603832669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,2560,0.006729777902364731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,2560,0.03124177787039015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,8192,0.10973778035905625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,2048,0.005817777580685086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,2048,0.026838221483760413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,6144,0.08627021974987453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,1536,0.00517511119445165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,7168,0.09742133484946357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,1536,0.023731556203630235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,1024,0.004389333228270213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,1024,0.019374221563339233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,5120,0.07607199748357137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,768,0.003991111285156674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,768,0.01755911111831665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,4096,0.06537777847713895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,512,0.003645333151022593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,512,0.016007110476493835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,3584,0.06018310785293579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,256,0.0032426667296224167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,256,0.014624888698259989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,3072,0.05414755476845635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,128,0.002983111060327954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,2560,0.049060444037119545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,128,0.013776000175211163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,64,0.002846222163902389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,64,0.013796444568369122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,256,32,0.0028817777832349143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,256,32,0.013579555683665805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,65536,0.054936889145109385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,65536,0.6814942359924316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,2048,0.04247466723124186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,16384,0.01146488885084788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,16384,0.18208889166514078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,1536,0.037083556254704796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,12288,0.012055111428101858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,12288,0.13823378086090088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,1024,0.031416889694001936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,10240,0.010642666783597736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,10240,0.11475377612643772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,768,0.03071111109521654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,8192,0.009308444129096137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,8192,0.08806488911310832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,512,0.028190222051408555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,7168,0.009118222528033787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,7168,0.07607199748357137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,256,0.028078221612506445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,6144,0.008360889222886827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,6144,0.06489511330922444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,256,128,0.027093332674768236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,5120,0.008344000412358178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,5120,0.05539377861552768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,65536,0.7738871044582791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,4096,0.007619555625650618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,4096,0.04589955674277412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,16384,0.211937771903144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,3584,0.007651555869314406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,3584,0.04139644569820828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,12288,0.1622568898730808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,3072,0.0068959999415609576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,10240,0.13593867090013292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,3072,0.03592177894380357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,2560,0.006200888918505774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,2560,0.03070311082734002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,8192,0.1098124451107449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,2048,0.005445333404673471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,2048,0.026452443665928308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,7168,0.09730399979485406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,1536,0.004823110997676849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,1536,0.023499555057949487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,6144,0.08606844478183323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,1024,0.00416711096962293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,1024,0.019253333409627277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,5120,0.0760515530904134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,768,0.0038053331275780997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,768,0.017629333668284945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,4096,0.06461155414581299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,512,0.003407111184464561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,512,0.01573333309756385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,3584,0.05953422519895765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,256,0.0030746666921509635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,3072,0.05326399869389004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,256,0.014476444986131458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,2560,0.04651733239491781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,128,0.0028719999310043124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,128,0.013742222554153867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,64,0.0027297778675953546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,64,0.01370488852262497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,128,32,0.002732444347606765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,128,32,0.013384888569513956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,65536,0.04543911086188423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,65536,0.6841742197672526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,16384,0.012679110798570844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,16384,0.18117154969109428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,12288,0.010321777727868822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,12288,0.1378764443927341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,10240,0.009344889058007134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,10240,0.11403911643558079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,8192,0.008344888687133789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,8192,0.08718133634991115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,7168,0.007931555310885111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,7168,0.07540000147289701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,6144,0.007377777662542131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,6144,0.06473777691523235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,5120,0.0070426662762959795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,5120,0.055042667521370776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,4096,0.006631999793979857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,4096,0.04580977890226576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,3584,0.006945778098371293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,3584,0.041265779071384005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,3072,0.006620444357395172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,3072,0.03618311219745212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,2560,0.006092444476154115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,2048,0.04022399915589227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,2560,0.03204800022972955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,2048,0.005372444374693765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,2048,0.02660622199376424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,1536,0.0046986668474144405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,1536,0.023574221465322707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,1024,0.004023111114899318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,1024,0.019229332605997723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,768,0.0037226668662495087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,768,0.017503110898865592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,512,0.003367111086845398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,512,0.015828443898095023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,256,0.0030337776988744736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,256,0.014434667097197639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,128,0.0028293333533737394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,128,0.013736888766288757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,64,0.0027075554761621687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,64,0.013645333548386892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,64,32,0.00272444449365139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,64,32,0.013337777720557319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,65536,0.04370577798949348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,16384,0.009510222408506606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,16384,0.18112266063690186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,65536,0.6825831201341418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,12288,0.008456000023418004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,10240,0.007797333101431529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,12288,0.13814844025505915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,8192,0.007140444384680853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,10240,0.11405955420600043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,7168,0.006634666687912411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,8192,0.08798488643434312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,6144,0.0064382221963670515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,7168,0.07607733541064791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,1536,0.03702755437956916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,5120,0.0069013333155049225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,6144,0.06513688961664836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,4096,0.0063484443558586975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,5120,0.05595733059777153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,4096,0.045787556303872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,3584,0.00675644435816341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,3072,0.006477333191368315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,3584,0.04110488957828946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,2560,0.006031999985376994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,3072,0.03743911120626662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,2048,0.005330666485759947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,2560,0.031856000423431396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,2048,0.02638133366902669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,1536,0.004661333229806689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,1024,0.003943110919660992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,1536,0.023511111736297607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,1024,0.031121777163611516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,1024,0.0192239996459749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,768,0.00361333332127995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,768,0.030777778890397813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,768,0.017398221625222098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,512,0.0032960000551409195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,128,0.0027955555253558685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,512,0.015838222371207345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,256,0.002973333415057924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,256,0.01457155578666263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,128,0.013697778185208639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,64,0.002628444383541743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,64,0.013599110974205865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,768,32,32,0.0026213334252436957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,768,32,32,0.01350311107105679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,16384,0.8733875486585828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,12288,0.6822417577107748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,16384,1.2409884134928386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,512,0.027879112296634253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,12288,1.1878471374511719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,10240,0.7467848989698621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,10240,0.5720844268798828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,8192,0.4374844498104519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,8192,0.6099617746141222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,7168,0.5459040006001791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,256,0.02785244584083557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,7168,0.38721688588460285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,6144,0.49646308686998153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,6144,0.33798578050401473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,12288,0.685914675394694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,16384,0.902223163180881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,5120,0.40136533313327366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,8192,0.4666764471266005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,5120,0.2710435655381945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,4096,0.33069509930080837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,4096,0.22595289018419054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,3584,0.19150400161743164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,7168,0.41355289353264707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,3584,0.2992044554816352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,10240,1.354347546895345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,6144,0.35539645618862575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,3072,0.2614622116088867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,3072,0.1656631098853217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,4096,0.2508311006757948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,2560,0.22269066174825033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,5120,0.3076355457305908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,2560,0.14049243927001953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,2048,0.1828577783372667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,2048,0.1178986628850301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,3584,0.22413155767652723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,1536,0.09764089186986287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,1536,0.13751199510362414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,1024,0.10032977660497029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,1024,0.07638933261235555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,2560,0.1732462247212728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,768,0.08157066504160564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,2048,0.1464444398880005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,1536,0.12017777231004502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,768,0.06576977835761176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,512,0.06099377738104927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,3072,0.3853315512339274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,1024,0.09486222267150879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,512,0.05643288956748115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,768,0.08328621917300753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,256,0.044138666656282216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,256,0.046725332736968994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,512,0.07149866554472181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,128,0.0362737774848938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,128,0.0425946679380205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,256,0.06885511345333524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,64,0.03211466802491082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,64,0.044940445158216685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,65536,128,0.06391377581490411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,65536,32,0.032976001501083374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,65536,32,0.05004622207747566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,16384,0.3023751046922472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,16384,0.27714310752020943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,12288,0.23322667015923393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,12288,0.2114551067352295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,10240,0.1927173270119561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,65536,1.3538204828898113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,10240,0.17849511570400664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,8192,0.17737421724531385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,8192,0.13900710476769343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,7168,0.14183021916283503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,12288,0.2439653343624539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,16384,0.3229271041022407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,7168,0.12262133757273357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,10240,0.20937777890099418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,6144,0.12296533584594727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,8192,0.17626221974690756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,6144,0.11447021696302627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,5120,0.10184178087446426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,7168,0.1554337739944458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,5120,0.09142933289210002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,4096,0.08364266819424099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,4096,0.07483288976881239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,6144,0.13832799593607584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,65536,4.009440104166667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,3584,0.07987555530336168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,3584,0.0680275559425354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,5120,0.11990311410692002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,4096,0.10161866744359334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,3072,0.060870223575168185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,2560,0.05607110924190945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,3072,0.3107395437028673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,3584,0.09311644236246745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,2560,0.05463555455207825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,3072,0.08397422234217326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,2048,0.046104888121287026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,2560,0.07597155703438653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,2048,0.09918044673071967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,1536,0.03773244553142124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,1536,0.03741511040263706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,2048,0.0650542245970832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,1024,0.027426666683620874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,1536,0.05630755424499512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,1024,0.058950225512186684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,768,0.022778666681713525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,1024,0.0480951103899214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,768,0.02698933415942722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,512,0.016928889685206942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,768,0.0437084436416626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,512,0.02424533334043291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,256,0.011997333003415002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,512,0.04050222370359633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,256,0.021334222621387903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,128,0.009379555781682333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,65536,1.2548756069607205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,256,0.03558133376969232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,128,0.019327110714382596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,64,0.007850666840871176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,16384,128,0.03403822249836392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,64,0.018944000204404194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,16384,32,0.008288000192907121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,16384,32,0.020420443680551317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,16384,0.23117510477701822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,16384,0.23226221402486166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,12288,0.17672355969746908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,65536,0.8859555456373426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,65536,0.940275510152181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,12288,0.17577955457899305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,10240,0.16324088308546278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,10240,0.14695910612742105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,8192,0.1255013280444675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,16384,0.27418221367730033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,8192,0.12769333521525064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,7168,0.10421688689125909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,12288,0.20790310700734457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,7168,0.10293244653277928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,6144,0.09608178006278144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,10240,0.17945155832502577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,6144,0.09159644444783528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,8192,0.14574044280582005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,5120,0.07871022489335802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,5120,0.07864799764421251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,6144,0.11912177668677436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,4096,0.06711466444863214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,4096,0.06366222434573703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,5120,0.10448533296585083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,7168,0.4741635322570801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,3584,0.06019466453128391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,3584,0.05667022201750013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,4096,0.08856622378031413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,3072,0.05236000153753492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,3584,0.08077510860231188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,2560,0.0439626673857371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,3072,0.15459556049770778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,2560,0.04753244585461087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,3072,0.07222311364279853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,2048,0.03626577721701728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,2560,0.06735199689865112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,2048,0.037662221325768366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,1536,0.029072887367672388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,1536,0.032264888286590576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,1024,0.02685600022474925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,1024,0.021143999364640977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,2048,0.1037786669201321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,1536,0.05021511183844673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,768,0.01697688963678148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,768,0.023665777511066858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,1024,0.04381155636575487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,256,0.009536888864305284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,65536,1.0487964418199327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,512,0.022486221459176805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,768,0.04087200098567539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,512,0.022809776994917128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,512,0.037857777542538114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,256,0.019518221418062847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,128,0.007439110842016008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,256,0.03327022327317132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,128,0.02146133283774058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,64,0.006312888943486744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,12288,128,0.030866665972603693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,12288,32,0.006598222172922558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,64,0.017484444710943434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,12288,32,0.018139556050300598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,16384,0.19238666693369547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,16384,0.2223688761393229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,12288,0.16469955444335938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,65536,0.8252942297193738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,12288,0.16767556137508818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,10240,0.13857155376010472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,8192,0.11000533236397637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,10240,0.1416648891237047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,8192,0.11220888296763103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,7168,0.09165600273344253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,65536,2.5092391967773438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,7168,0.09992355770534939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,6144,0.07937866449356079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,6144,0.08825777636633979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,768,128,128,0.027066666218969557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,5120,0.06697599755393134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,5120,0.0759119987487793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,12288,0.20431466897328696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,10240,0.1668684482574463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,4096,0.0573128859202067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,4096,0.06208088662889269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,8192,0.14566577805413142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,3584,0.050494223833084106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,3584,0.055919110774993896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,6144,0.11344888475206162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,3072,0.04410755634307861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,7168,0.12761244508955213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,16384,0.2559751139746772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,3072,0.04973155591222975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,2560,0.03858933183881972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,65536,0.9616133371988932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,2560,0.04373866650793287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,2048,0.03243022163709005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,2048,0.03610310951868693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,1536,0.025988444685935974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,1536,0.03131999903255039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,1024,0.01834844383928511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,1024,0.026183999247021143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,5120,0.09888444344202678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,768,0.015117333994971381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,768,0.023450666003757056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,3584,0.07700710826449923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,512,0.012092444631788464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,512,0.021643555826610986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,2560,0.06384888622495863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,256,0.009103111094898647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,256,0.01923288901646932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,1536,0.04842666784922282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,128,0.0069288888739215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,128,0.01716888944307963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,4096,0.08370311392678155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,64,0.006167999986145232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,64,0.017495110630989075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,10240,32,0.0063662222690052455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,10240,32,0.01774311065673828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,65536,0.6490951114230686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,2048,0.05572266711129082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,3072,0.06982133123609754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,16384,0.15044000413682726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,65536,0.7273573345608182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,16384,0.18337066968282065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,12288,0.11750400066375732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,12288,0.13905333148108587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,16384,0.23699821366204155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,10240,0.09931111335754395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,10240,0.11620710955725776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,12288,0.1888951063156128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,8192,0.09645777940750122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,8192,0.09705244170294867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,10240,0.15059378412034777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,65536,0.852367983924018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,7168,0.07288089063432482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,7168,0.08385155598322551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,8192,0.1323217815823025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,6144,0.06453244553671943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,7168,0.11335555712381999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,6144,0.07523022095362346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,5120,0.054788443777296275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,5120,0.06473955843183729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,6144,0.10222399897045559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,4096,0.04337955514589945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,4096,0.052236444420284696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,5120,0.08963821993933783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,3584,0.040594667196273804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,3584,0.0473306675752004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,4096,0.07694666915469699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,3072,0.03492355677816603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,3584,0.07031466563542683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,1024,0.0428711109691196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,3072,0.04253510965241326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,2560,0.03027555677625868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,2560,0.040384888648986816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,3072,0.06404444244172838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,2048,0.025452444950739544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,2048,0.031077331966824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,2560,0.060270223352644176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,1536,0.020733333296246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,2048,0.05017688870429993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,1536,0.02646933330429925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,1024,0.015024888846609326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,1024,0.022872888379626807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,1536,0.04488977789878845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,768,0.012813333008024426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,1024,0.03912711143493652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,768,0.02102222210831112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,512,0.010241777532630498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,768,0.03836888737148709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,512,0.01868799991077847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,256,0.007703999678293864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,512,0.03468177715937296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,128,0.006080889039569431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,256,0.017280000779363845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,256,0.031026668018764917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,128,0.01575377749072181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,64,0.005665777872006099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,8192,128,0.030181331766976252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,64,0.015941333439615037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,8192,32,0.005797333187527127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,8192,32,0.01609511176745097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,16384,0.15190222528245714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,65536,0.5950586530897353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,65536,0.712774223751492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,16384,0.21089333958095977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,12288,0.1659031046761407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,12288,0.159844438234965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,16384,0.22789600160386828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,10240,0.09581866529252793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,10240,0.11130311754014756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,768,0.039192001024881996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,12288,0.2497048907809787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,8192,0.0783626635869344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,8192,0.1325920025507609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,65536,0.8554293314615885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,10240,0.1487306621339586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,7168,0.11559644010331895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,7168,0.08145866791407268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,8192,0.1221662229961819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,6144,0.0813608898056878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,7168,0.11147467295328777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,6144,0.07532888650894165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,5120,0.05503377649519178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,5120,0.07442666424645318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,6144,0.09980444113413493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,4096,0.039664887719684176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,4096,0.051070223251978554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,5120,0.08809155225753784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,3584,0.04624799887339274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,3584,0.03699644406636556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,4096,0.08958577447467381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,3072,0.031034668286641438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,3072,0.041577776273091636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,3584,0.06854844755596586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,2560,0.03545244534810384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,3072,0.06368622514936659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,2560,0.03923199905289544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,2048,0.034656001461876765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,2048,0.022351110974947613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,1536,0.01738666660255856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,2560,0.059883554776509605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,1536,0.02858577834235297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,2048,0.05059022373623318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,1024,0.013047110703256396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,1024,0.02293511066171858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,1536,0.044657776753107704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,1024,0.04103644357787238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,768,0.01054844425784217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,768,0.020933333370420668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,512,0.008254222571849823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,512,0.020230222079488967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,768,0.03793688946300083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,256,0.006251555350091722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,256,0.016559999850061204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,512,0.0354693333307902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,128,0.0052577778697013855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,256,0.03123377760251363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,128,0.01570044457912445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,64,0.015836444165971544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,32,0.005184000151024925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,7168,64,0.004946666873163647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,7168,128,0.030138668086793687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,7168,32,0.01590044465329912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,16384,0.11491200659010146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,65536,0.44628535376654727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,16384,0.1772480010986328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,12288,0.10452089044782852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,12288,0.13258666462368435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,65536,0.9383298026190864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,10240,0.07886399825414021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,16384,0.20351644357045492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,10240,0.15966222021314833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,8192,0.06169866853290134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,12288,0.1577226718266805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,8192,0.12221954928504096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,7168,0.06003644731309679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,10240,0.13401244746314153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,7168,0.09610311190287273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,8192,0.11043110820982192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,6144,0.04978133241335551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,7168,0.10223999950620864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,6144,0.07846489217546251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,5120,0.04095200035307143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,5120,0.06215200159284803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,6144,0.09059200021955703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,4096,0.034068445364634194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,4096,0.05065600077311198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,65536,0.7565075556437174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,5120,0.09966311189863418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,3584,0.030157334274715845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,4096,0.06837066676881579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,3584,0.04569244384765625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,3072,0.03528711199760437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,3072,0.04268533322546217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,3584,0.06378489070468478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,3072,0.05933155616124471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,2560,0.02403822210099962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,2048,0.019387554791238572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,2560,0.045273777511384755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,2048,0.03077955709563361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,2560,0.05505422088834974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,2048,0.04686933424737719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,1536,0.027669333749347266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,1536,0.015487111277050443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,1024,0.01202222208182017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,1536,0.0417208903365665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,1024,0.022809776994917128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,768,0.01036800030204985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,1024,0.03670933180385166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,512,0.008382221890820397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,768,0.0206977774699529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,768,0.035770667923821345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,256,0.0064088888466358185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,512,0.018349332941903006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,512,0.032608889871173434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,256,0.01683999929163191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,128,0.005149333427349727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,256,0.029425776667065088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,128,0.015607110328144498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,64,0.004652444273233414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,128,0.03040800160831875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,6144,128,0.029205332199732464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,64,0.015809777710172866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,6144,32,0.004863111095296012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,6144,32,0.015765332513385348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,16384,0.10231377681096394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,65536,0.43874311447143555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,16384,0.1766453319125705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,12288,0.08609422047932942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,65536,0.6925493346320258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,12288,0.13054933812883165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,16384,0.19651467270321318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,10240,0.06748000118467543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,10240,0.10849510961108738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,12288,0.15002311600579157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,8192,0.05383377936151293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,8192,0.08845688899358113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,7168,0.05078577664163378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,256,0.03307377629809909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,7168,0.08425244357850815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,10240,0.12962578402625188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,8192,0.10821155707041423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,6144,0.04450755649142795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,7168,0.09792266951666938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,6144,0.07051377826266818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,5120,0.03849955399831136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,5120,0.06089511182573107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,6144,0.10840800073411729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,4096,0.0314302212662167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,65536,0.7133768929375542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,4096,0.06715289089414808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,5120,0.07923199733098348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,4096,0.05000266763899061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,3072,0.04155644443300035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,3584,0.061956445376078285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,3584,0.027967111931906805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,3584,0.04952444301711189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,3072,0.024466666910383437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,2560,0.021197333931922913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,2560,0.03611466619703505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,3072,0.058006220393710665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,2048,0.017265778448846605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,2048,0.0447599987188975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,2560,0.053731554084353976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,2048,0.0307626658015781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,1536,0.014667555689811707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,1536,0.02594311038653056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,1024,0.011524444652928246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,768,0.019314666589101154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,1536,0.04043822156058417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,1024,0.02200355629126231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,768,0.00997155573632982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,256,0.0063564446237352155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,1024,0.036292443672815956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,512,0.008271111382378472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,768,0.03436977664629618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,512,0.01793244481086731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,512,0.031097776359981958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,256,0.016255999604860943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,128,0.005056000004212062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,64,0.015572445260153877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,256,0.029320889049106177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,128,0.015351111690203348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,64,0.0046817776229646476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,5120,128,0.02858933475282457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,65536,0.3475182321336534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,5120,32,0.004736888739797804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,5120,32,0.015424888994958667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,16384,0.0858888891008165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,12288,0.06081777811050415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,16384,0.13957866032918295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,65536,0.545721795823839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,12288,0.1028711133533054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,10240,0.05238222082455953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,10240,0.08604266908433701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,8192,0.04411822226312426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,8192,0.07659377654393514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,7168,0.0384808878103892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,7168,0.06919911172654894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,6144,0.038167999850379095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,6144,0.06016088856591118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,10240,512,0.03608355588383145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,5120,0.029394666353861492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,5120,0.050605333513683744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,16384,0.1857759952545166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,4096,0.023769777682092454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,4096,0.040656887822681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,12288,0.14088533984290227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,10240,0.12073866526285808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,3584,0.021193777521451313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,3584,0.03812444541189406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,3072,0.01886311173439026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,8192,0.1006559994485643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,65536,0.6759351094563802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,3072,0.03424800104565091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,7168,0.0920906662940979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,2560,0.015953777564896476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,2560,0.030304888884226482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,6144,0.08276977803972033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,2048,0.014076444837782117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,2048,0.026054221722814772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,1536,0.011341333389282227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,1536,0.023240889112154644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,1024,0.009122666385438707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,1024,0.01981244484583537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,5120,0.07335022423002455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,768,0.007641777396202087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,768,0.017324444320466783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,3584,0.05834844377305773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,512,0.006127111199829314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,4096,0.06252177556355794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,512,0.016113777955373127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,256,0.004968889057636261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,256,0.015154666370815702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,3072,0.05381688806745741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,128,0.004359111189842224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,128,0.014375999569892883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,2560,0.04964977833959791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,1536,0.03840711050563388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,64,0.003976888954639435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,64,0.014412444498803882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,4096,32,0.0042257776690853965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,2048,0.0431182218922509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,4096,32,0.014156444205178155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,65536,0.26605065663655597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,16384,0.07206222083833483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,65536,0.5464906692504883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,1024,0.03481422199143304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,16384,0.1380479998058743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,12288,0.05526933405134413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,12288,0.10076355271869236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,10240,0.04787733488612705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,10240,0.08378310998280843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,768,0.03242222136921353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,8192,0.0374017788304223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,8192,0.06980533070034452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,256,0.028393778536054824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,7168,0.03440533412827386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,512,0.031150221824645996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,7168,0.06266578038533528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,4096,128,0.028049776951471966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,6144,0.03125955661137899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,6144,0.054959999190436475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,5120,0.025437333517604407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,5120,0.04890755481190152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,16384,0.19832355446285674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,4096,0.02404888967672984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,4096,0.041737778319252856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,65536,0.6938444243537055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,3584,0.019153777096006606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,12288,0.14053955343034533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,3584,0.03723466727468703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,10240,0.11952355172899033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,3072,0.016840888394249808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,3072,0.033167110549079046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,2560,0.014563555518786112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,2560,0.029776000314288672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,8192,0.09950755702124701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,2048,0.012631110846996307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,2048,0.025772444076008264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,7168,0.09082666370603774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,6144,0.08117333385679457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,1536,0.01070577816830741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,1536,0.022954665952258643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,1024,0.008375110725561777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,1024,0.01975822283162011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,5120,0.07169066535101996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,768,0.0069573331210348345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,768,0.017228444417317707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,4096,0.062353776560889355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,512,0.005679111099905438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,512,0.016040000650617812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,3584,0.0582559969690111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,256,0.0047155556579430895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,3072,0.05373866690529717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,256,0.015058666467666626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,2560,0.05012889040840996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,128,0.0041511112617121804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,128,0.014295111099878946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,64,0.0037457777394188773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,64,0.014315555493036905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3584,32,0.003907555507289039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3584,32,0.014155555102560254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,2048,0.04382488793796963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,65536,0.23994665675693086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,65536,0.5212151209513346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,16384,0.062421335114373096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,16384,0.13717244731055364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,1536,0.03903022077348497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,12288,0.05008800162209404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,12288,0.10084355539745754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,768,0.03159200151761373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,10240,0.05113244387838575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,10240,0.0835591091050042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,1024,0.033891555335786604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,8192,0.03354933195643955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,8192,0.06832889053556654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,256,0.02824266751607259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,7168,0.030844443374209937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,128,0.028322666883468628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3584,512,0.029920889271630183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,7168,0.06196622053782145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,6144,0.02616888946957058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,6144,0.0546648899714152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,5120,0.02776088813940684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,5120,0.04789066645834181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,65536,0.6643502447340224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,4096,0.019859555694792006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,4096,0.039078222380744085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,16384,0.1834862232208252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,3584,0.01685688893000285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,3584,0.03609777821434869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,12288,0.13899021678500706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,3072,0.014908444550302295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,3072,0.03292355603641934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,10240,0.11881777975294326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,8192,0.09934400187598334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,2560,0.013129777378506131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,2560,0.029907554388046265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,2048,0.01126399967405531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,2048,0.026029333472251892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,7168,0.09050310982598199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,1536,0.00942844483587477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,1536,0.02283199959331089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,6144,0.08153688907623291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,5120,0.07239644394980536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,1024,0.0074444446298811175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,1024,0.01849688920709822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,768,0.006191111273235745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,768,0.017127111554145813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,3584,0.0580728915002611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,512,0.005400889035728242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,512,0.016207999653286405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,4096,0.0620044469833374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,256,0.004474666797452503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,256,0.014998222390810648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,768,0.0328800015979343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,128,0.003955555458863576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,128,0.014214222629865011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,2560,0.04877510997984144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,64,0.003698666476541095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,64,0.014315555493036905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,3072,32,0.003855111284388436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,3072,32,0.014193778236707052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,65536,0.209678226047092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,3072,0.05280444357130262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,65536,0.5198657777574327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,1536,0.03882311118973626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,16384,0.06316444608900282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,16384,0.13638844754960802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,12288,0.05282488796446058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,12288,0.09966666830910577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,2048,0.043154666821161904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,10240,0.048044443130493164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,10240,0.08289777570300631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,1024,0.03382488754060533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,8192,0.0298977792263031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,8192,0.06780178017086454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,256,0.027767111857732136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,7168,0.03209599852561951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,7168,0.06047822369469536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,512,0.029932445949978296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,6144,0.02852355440457662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,6144,0.05421955717934502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,3072,128,0.027752000424597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,5120,0.026123555170165166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,5120,0.04748444424735176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,65536,0.6465688811408149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,4096,0.0163111107216941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,4096,0.03896355628967285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,16384,0.17998222510019937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,3584,0.016335111525323655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,3584,0.03594933284653558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,12288,0.13643288612365723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,3072,0.015987555185953777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,3072,0.03274755676587423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,10240,0.11759732829199897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,2560,0.013415999710559845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,2560,0.029747555653254192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,8192,0.09812533193164402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,2048,0.010393777655230628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,2048,0.025646222962273493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,7168,0.08911910984251235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,1536,0.009256000320116678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,6144,0.08110577530331083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,1536,0.022427555587556627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,1024,0.006619555668698416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,1024,0.01830933325820499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,5120,0.07214133607016669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,768,0.006007110906971826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,768,0.017128000656763714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,3584,0.05720622009701199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,512,0.005024000174469418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,512,0.0163102216190762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,2560,0.049208889404932656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,256,0.004234666625658671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,256,0.014886221951908536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,4096,0.061582220925225147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,128,0.0037662221325768363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,128,0.014223110344674854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,3072,0.05117333266470167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,64,0.003528000166018804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,64,0.014151111245155334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2560,32,0.0037191112836201987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2560,32,0.014005333185195923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,65536,0.16310400433010525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,65536,0.5176711082458496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,2048,0.042113777663972646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,16384,0.049978666835361056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,16384,0.13605688677893743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,768,0.030961778428819444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,12288,0.050256000624762646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,12288,0.0988960001203749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,1536,0.03800266649987962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,10240,0.043291555510626904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,10240,0.0813244448767768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,1024,0.03419466813405355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,8192,0.03176800078815884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,8192,0.06715377834108141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,128,0.027689778142505225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,7168,0.02868888775507609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,7168,0.06090133057700264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,256,0.02743822170628442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,6144,0.026791999737421673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,6144,0.052720887793434985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,65536,0.6078480084737142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,16384,0.1676044464111328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,5120,0.022104889154434204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,5120,0.04678933487998115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,4096,0.017117333081033494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,4096,0.03854755560557047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2560,512,0.029354665014478896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,3584,0.016421332955360413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,3584,0.03568799959288703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,12288,0.12588977813720703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,3072,0.014175110393100314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,3072,0.032165331972969904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,10240,0.10771110985014175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,2560,0.012350222302807702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,2560,0.029620445436901514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,8192,0.08966133329603408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,2048,0.010911111202504901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,2048,0.025166221790843542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,7168,0.08247111241022746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,1536,0.008837333156002892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,1536,0.021856000026067097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,6144,0.07458844449785021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,1024,0.006714666469229593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,1024,0.01832355558872223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,5120,0.06656710969077216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,768,0.005576000031497743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,768,0.017049777838918898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,4096,0.057512885994381376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,512,0.004621333132187526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,512,0.015860444969601102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,3584,0.05347999930381775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,256,0.004002666721741359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,256,0.015125332607163323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,3072,0.04890311095449659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,128,0.0035546666218174826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,128,0.014155555102560254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,2560,0.04650222261746725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,64,0.0033484444850020935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,64,0.0140364451540841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,2048,32,0.003463111196955045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,2048,32,0.013952000273598565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,65536,0.125709335009257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,65536,0.5101555718315972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,2048,0.04065066576004028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,16384,0.040814220905303955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,16384,0.13490400049421522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,1536,0.036702222294277616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,12288,0.039697776238123574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,12288,0.09728888670603435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,1024,0.031275557147132024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,10240,0.03719644414054023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,10240,0.08018844657474093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,512,0.02861600120862325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,8192,0.026849778162108526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,8192,0.06544088655047946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,768,0.030333333545260962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,7168,0.02425866656833225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,7168,0.059564444753858775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,256,0.02720444401105245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,2048,128,0.027274666561020747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,6144,0.021570665968788996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,6144,0.05239644315507677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,5120,0.01998933321899838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,5120,0.046354668007956616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,65536,0.6181591351826986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,16384,0.17061155372195771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,4096,0.014588443769348992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,4096,0.038314667012956403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,3584,0.01385155568520228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,3584,0.034557332595189415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,12288,0.12943022780948216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,3072,0.012808000048001608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,3072,0.03220444586541917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,10240,0.10999911361270481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,2560,0.010690666735172272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,2560,0.029346668057971533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,8192,0.09209244118796454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,2048,0.009296889106432596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,2048,0.02423288921515147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,7168,0.08417955372068618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,1536,0.0076044441925154785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,1536,0.021361778179804485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,6144,0.07575733131832547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,1024,0.005679111099905438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,1024,0.018185777796639335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,5120,0.06727821959389581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,768,0.004938666605287128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,768,0.016939555605252583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,4096,0.057277335060967334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,512,0.0041404445138242514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,512,0.015876443849669564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,3584,0.05237955517239041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,256,0.003616888903909259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,256,0.01477866702609592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,3072,0.04822755522198147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,128,0.003314666656984223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,128,0.014158222410413953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,2560,0.04558222161398994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,64,0.003091555502679613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,64,0.013873777455753751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1536,32,0.0032462223122517266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,768,0.029311110575993855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1536,32,0.013833777772055732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,65536,0.08939555618498062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,65536,0.5089022318522135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,16384,0.03135111265712314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,2048,0.03997777899106344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,1536,0.035939554373423256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,16384,0.13422400421566433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,12288,0.028927998410330877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,12288,0.09626133574379815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,10240,0.03009866674741109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,10240,0.07989688714345296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,1024,0.03036622206370036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,8192,0.021198223034540813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,8192,0.06502311097251044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,256,0.027061333258946735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,7168,0.020079111059506733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,7168,0.05885244740380181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,512,0.028231110837724473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,6144,0.015352000792821249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,6144,0.051572445366117686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1536,128,0.02725155486000909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,5120,0.015925332903862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,5120,0.04551644457711113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,65536,0.5720257759094238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,4096,0.01168444421556261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,4096,0.03768444392416213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,16384,0.15689955817328557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,3584,0.011024889018800525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,12288,0.11900088522169326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,3584,0.034767998589409724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,3072,0.009764444496896531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,10240,0.10158577892515396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,3072,0.03138400117556254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,2560,0.007756444315115611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,2560,0.027808000644048054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,8192,0.08516977892981635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,2048,0.006587555425034628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,2048,0.023660444551044043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,7168,0.07755466964509752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,1536,0.005620444400442972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,1536,0.020765332712067496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,6144,0.0705288913514879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,1024,0.004699555536111196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,1024,0.017900443739361234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,5120,0.06278133392333984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,768,0.00434044458799892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,768,0.016703110602166917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,4096,0.05385955505900913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,512,0.0039057777159743835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,512,0.01552177800072564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,3072,0.04601688848601448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,3584,0.04942311180962456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,256,0.0034506666577524612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,256,0.014712888333532544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,2560,0.04379466507169935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,128,0.003240888938307762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,128,0.01384622189733717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,2048,0.038662221696641706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,64,0.003055111194650332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,64,0.01366488883892695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,1024,32,0.003145777723855443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,1024,32,0.013655111193656921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,65536,0.07428533501095243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,65536,0.5099982155693902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,16384,0.026463111241658527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,16384,0.1344506608115302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,1536,0.03292799989382426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,12288,0.025026667449209426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,12288,0.09670488701926337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,1024,0.0295413335164388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,10240,0.02237866653336419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,10240,0.07863466607199775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,768,0.028852442900339764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,8192,0.01661688917213016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,8192,0.06540355417463514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,512,0.027541332774692114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,7168,0.014979556202888489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,7168,0.05894311269124349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,256,0.026742221580611333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,6144,0.013516444298956128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,1024,128,0.026783999469545152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,6144,0.051522665553622775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,5120,0.013873777455753751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,5120,0.045082665152019925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,65536,0.5604462093777126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,16384,0.15463022391001383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,4096,0.010527999864684211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,4096,0.03740977909829881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,3584,0.010148444109492833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,3584,0.03464533223046197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,12288,0.11522311634487575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,3072,0.008272888759771982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,3072,0.030908445517222088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,10240,0.09904177983601888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,2560,0.007768889268239339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,2560,0.027448000179396734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,8192,0.08456622229682074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,2048,0.006870222174459034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,2048,0.023611555496851604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,7168,0.07646577888064914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,1536,0.006031111296680238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,1536,0.020488000578350492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,5120,0.06157333321041531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,6144,0.06907288895712958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,1024,0.004828444371620814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,1024,0.017779555585649278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,768,0.004117333226733738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,768,0.016540444559521146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,4096,0.05225244495603773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,512,0.003690666622585721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,512,0.015285332997639975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,3584,0.04907733201980591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,256,0.0033324443631701996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,256,0.014709333578745524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,3072,0.04543022314707438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,128,0.0030586665703190696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,128,0.013805333111021252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,2560,0.043191999197006226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,64,0.0028657778683635923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,64,0.013560000393125745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,768,32,0.0029519999192820657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,768,32,0.013575110998418597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,65536,0.05743733379575941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,65536,0.5073119799296061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,2048,0.036229332288106285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,16384,0.020249777370029025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,16384,0.1346773306528727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,1536,0.03291644321547614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,12288,0.01800355480776893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,1024,0.029622223642137315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,768,0.028597331709331934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,12288,0.09679200251897176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,10240,0.01833599971400367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,10240,0.0793608890639411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,8192,0.013820444544156393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,8192,0.06536444690492418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,512,0.027634667025672063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,7168,0.013518222504191928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,7168,0.05837066968282064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,256,0.02660622199376424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,6144,0.014262222581439547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,6144,0.051346666283077665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,768,128,0.02642044425010681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,5120,0.012049777640236748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,5120,0.04446755515204536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,65536,0.5492497550116645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,4096,0.009474666582213508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,4096,0.03700088792377048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,16384,0.1524897813796997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,3584,0.008813333180215623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,3584,0.03456444541613261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,10240,0.09722755352656047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,12288,0.11418844593895806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,3072,0.007939555578761632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,3072,0.030254221624798242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,2560,0.007159111400445302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,8192,0.08247822523117065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,2560,0.027619555592536926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,2048,0.006467555546098285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,7168,0.07535555627610949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,2048,0.02345955537425147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,1536,0.005629333357016246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,6144,0.06821244292789035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,1536,0.02053511142730713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,1024,0.004671999977694618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,1024,0.017924444542990792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,5120,0.06007555458280775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,768,0.0042071110672420925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,768,0.01645066671901279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,4096,0.051628443929884166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,512,0.00370755543311437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,512,0.015246222416559855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,3584,0.048242668310801186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,256,0.0033004445334275565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,256,0.014527999692493014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,3072,0.044053332673178784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,128,0.003015999992688497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,128,0.013735999663670858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,2560,0.04256711072391934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,64,0.0028924445311228433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,64,0.013665777941544851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,512,32,0.0029288888391521242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,512,32,0.0135057775510682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,65536,0.050423112180497914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,65536,0.5101057688395182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,2048,0.0378115541405148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,16384,0.013717333475748697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,16384,0.1344577736324734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,1536,0.03204444381925795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,12288,0.011791110866599612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,1024,0.029241777128643457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,12288,0.0974044468667772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,10240,0.010500444306267632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,10240,0.08023199770185682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,768,0.028171555863486394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,8192,0.009208888643317753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,8192,0.0640711122088962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,512,0.027777777777777776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,7168,0.011163555913501315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,7168,0.057638221316867404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,256,0.02624888883696662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,6144,0.00884622252649731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,6144,0.050865779320398964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,512,128,0.026591110560629103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,5120,0.008940444224410588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,5120,0.04355466696951124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,65536,0.5485031339857314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,4096,0.008188444707128737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,4096,0.03646488984425863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,16384,0.1506462229622735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,3584,0.007731555236710443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,3584,0.03288711110750834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,12288,0.11376444498697917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,3072,0.007263999846246507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,10240,0.09672711292902629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,3072,0.029398222764333088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,2560,0.006335999816656113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,2560,0.026573333475324843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,8192,0.0812915563583374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,2048,0.005596444424655702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,2048,0.022991110881169636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,7168,0.07435999976264106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,1536,0.004858666823969947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,1536,0.020247111717859905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,6144,0.06740533643298678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,1024,0.0041831110914548235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,1024,0.017875555488798354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,5120,0.05889955494138929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,768,0.003906666818592283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,768,0.01646488904953003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,4096,0.0510773327615526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,512,0.003544888976547453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,512,0.015221332510312399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,3584,0.0472906662358178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,256,0.003136000078585413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,256,0.014175110393100314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,3072,0.044575111733542554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,128,0.0029519999192820657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,2560,0.03979288869433933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,128,0.013815999858909182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,64,0.0028248888750871024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,64,0.013626666532622443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,256,32,0.002805333377586471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,256,32,0.013579555683665805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,65536,0.04119644562403361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,65536,0.5109599961174859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,2048,0.035079111655553184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,16384,0.011063999599880643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,16384,0.1341093381245931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,1536,0.031521777311960854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,12288,0.009544000029563904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,12288,0.09653244415918986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,1024,0.02899555696381463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,10240,0.01019200020366245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,10240,0.07973866992526583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,768,0.028183112541834515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,512,0.0270142224099901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,8192,0.008298666940795051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,8192,0.06440444125069512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,7168,0.007817777494589487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,7168,0.05786133474773831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,256,0.02611644400490655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,6144,0.008747555315494537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,6144,0.05033333433998955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,256,128,0.026501332720120747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,5120,0.008081778056091731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,5120,0.04372355673048231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,65536,0.5475031005011665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,4096,0.00742222203148736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,4096,0.03637511200375027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,16384,0.1503804392284817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,3584,0.007070221834712558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,3584,0.032207111517588295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,12288,0.11322222815619574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,3072,0.006718222051858902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,3072,0.028647998968760174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,10240,0.09753866990407307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,2560,0.006226666685607698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,2560,0.02645688917901781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,8192,0.0821031133333842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,2048,0.005478222337034013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,7168,0.07500088877148099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,2048,0.02311822275320689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,1536,0.0047893333766195504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,1536,0.020253333780500624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,6144,0.06708266337712605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,1024,0.0040942223535643685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,1024,0.01756977703836229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,5120,0.06007466713587443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,768,0.0037573331760035623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,768,0.01649688846535153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,4096,0.05089866783883837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,512,0.003399111123548614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,512,0.015274667077594332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,3584,0.045704887972937695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,256,0.003055111194650332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,256,0.014209777116775513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,3072,0.04313155677583483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,128,0.0028284444577164124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,128,0.013635555903116861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,2560,0.039220445685916476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,64,0.0027751111321979095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,64,0.01351466692156262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,128,32,0.0028888889484935333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,128,32,0.013272888958454132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,65536,0.033744888173209295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,65536,0.5110061963399252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,16384,0.009789333575301701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,16384,0.1342408921983507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,12288,0.008195555872387355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,12288,0.09691022502051459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,10240,0.0075528886583116316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,10240,0.07921244700749715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,8192,0.006984000404675801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,8192,0.0645973351266649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,7168,0.00676711110605134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,7168,0.057760000228881836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,6144,0.006496888895829518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,2048,0.034824887911478676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,6144,0.050494223833084106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,5120,0.006811555474996567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,4096,0.006441777778996362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,5120,0.04362755682733324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,4096,0.034874667723973594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,3584,0.006873777757088344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,1536,0.031927112076017596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,3584,0.03180799881617228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,3072,0.006531555619504716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,2560,0.005919999960396025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,3072,0.028764443265067205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,2560,0.02648799949222141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,2048,0.0052897776994440295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,1024,0.028767112228605483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,1536,0.0046693334976832075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,2048,0.02306488818592495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,1536,0.020096888144810993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,768,0.028176890479193792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,1024,0.003993777765168084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,1024,0.01757866640885671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,768,0.003683555457327101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,768,0.0162462227874332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,512,0.0033191111352708605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,512,0.015037332971890768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,256,0.002965333354141977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,128,0.002784888984428512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,256,0.014231999715169271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,128,0.013607111242082385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,64,0.0026364444444576898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,64,0.013417777915795645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,64,32,0.0026568888376156488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,64,32,0.013267555170589022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,16384,0.007792000141408708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,65536,0.033709334002600774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,12288,0.007115555306275685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,16384,0.13402488496568468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,10240,0.006803555621041193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,65536,0.5079181989034017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,12288,0.09694666994942559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,8192,0.006680888848172293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,10240,0.07886577977074517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,512,0.027561777167850073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,7168,0.006591111007663939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,8192,0.0640044477250841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,6144,0.006364444477690591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,7168,0.05769511063893636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,6144,0.05043199989530775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,5120,0.006695999867386288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,5120,0.04215733210245768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,4096,0.006342222293217977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,4096,0.034689777427249484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,3584,0.00685422205262714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,3072,0.006405333264006509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,3584,0.03229422370592753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,3072,0.02877955635388692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,2560,0.005970666805903117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,2048,0.005295111073387994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,2560,0.02644355595111847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,2048,0.023132445083724126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,1536,0.004685333205593957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,768,0.003649777836269803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,1536,0.020049777295854356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,1024,0.003969777789380816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,1024,0.01779022150569492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,768,0.016342222690582275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,512,0.003293333368168937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,256,0.0029644444584846497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,512,0.015230221880806817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,256,0.014095111025704278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,128,0.002776888923512565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,128,0.013584000368913015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,64,0.0026506665680143568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,64,0.013332444760534497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,512,32,32,0.0026400000270869997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,512,32,32,0.013191999660597907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,256,0.026446221603287592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,16384,1.0136817296346028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,12288,0.8275955518086752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,10240,0.6561822361416286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,12288,0.8867457707722982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,16384,1.19013426038954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,8192,0.49562401241726345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,10240,0.7430231306287977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,8192,0.5906942155626085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,7168,0.4533279736836751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,7168,0.5161546601189507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,6144,0.3886871072981093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,6144,0.4447973569234212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,5120,0.31991200976901585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,5120,0.36899201075236004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,512,128,128,0.02639644510216183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,4096,0.25587111049228245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,8192,0.36586488617791074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,4096,0.2971839904785156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,12288,0.5331777996487087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,3584,0.23061601320902506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,7168,0.33058667182922363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,3584,0.2601777712504069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,3072,0.20439199606577554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,6144,0.29343199729919434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,16384,0.686755551232232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,3072,0.2246613237592909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,10240,1.0900248421563041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,2560,0.17376355330149332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,2560,0.1882693370183309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,2048,0.1783191098107232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,2048,0.1522106726964315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,1536,0.1353457768758138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,1536,0.11600800355275472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,1024,0.07866488562689887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,1024,0.08266399966345893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,5120,0.24373955196804473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,768,0.06625955634646945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,768,0.06502933634652032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,768,0.07369066609276666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,512,0.05053955647680494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,512,0.05134222242567274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,2560,0.1416915522681342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,256,0.039429333474900984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,256,0.041364444626702204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,3584,0.18939199712541369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,128,0.030264000097910564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,128,0.03784266776508755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,1536,0.10666933324601914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,64,0.02793688906563653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,64,0.04475466741455925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,65536,32,0.029678222205903795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,65536,32,0.04742666747834948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,4096,0.2033511135313246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,3072,0.16085510783725315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,2048,0.11999555428822835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,65536,1.260430230034722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,16384,0.25170932875739205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,65536,1.3724230660332575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,12288,0.1841191053390503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,16384,0.3412764337327745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,12288,0.253711117638482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,1024,0.08127378092871772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,10240,0.1519333389070299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,10240,0.21279466152191162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,8192,0.12152889039781357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,8192,0.1710560056898329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,128,0.04923911227120293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,7168,0.11469066143035889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,7168,0.15097510814666748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,256,0.055941330062018506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,6144,0.10092355807622273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,6144,0.13255111376444498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,65536,512,0.06394222047593859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,5120,0.08734577894210815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,5120,0.11072622405158149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,16384,0.24629510773552787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,4096,0.06843555635876127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,4096,0.0911324421564738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,12288,0.20901689264509413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,3584,0.06880444288253784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,3584,0.09795555803510879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,65536,0.9601928922865125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,3072,0.053822222683164805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,3072,0.07130222188101874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,10240,0.16748355494605172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,2560,0.04407199886110094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,2560,0.060324443711174854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,8192,0.1378400060865614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,2048,0.036554667684766985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,2048,0.0510444442431132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,7168,0.1262764400906033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,1536,0.030262221892674763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,6144,0.11322311560312907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,1536,0.04180888997183906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,1024,0.021244444780879553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,1024,0.03280977739228143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,5120,0.09832355711195205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,768,0.017528888252046373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,768,0.02758933272626665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,3584,0.08771644698248969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,512,0.01573955516020457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,512,0.026168000366952684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,4096,0.08389244476954143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,256,0.010002666877375709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,256,0.021366222037209406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,2560,0.06323733594682482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,128,0.007771555748250749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,128,0.01865244408448537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,3072,0.07195378012127347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,64,0.00656444455186526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,64,0.018799111247062683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,16384,32,0.006763555523422029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,16384,32,0.019848000672128465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,65536,0.7381280263264974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,2048,0.05525955557823181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,65536,1.0986088646782768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,16384,0.18712355030907524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,16384,0.2693706618414985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,768,0.04026577870051066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,12288,0.13574311468336317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,12288,0.20031111770206025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,1536,0.05033599999215868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,10240,0.11327466699812148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,10240,0.16871999369727242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,1024,0.04277866747644213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,8192,0.0977591143714057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,8192,0.13559199704064265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,128,0.02910311023394267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,7168,0.07985155449973212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,7168,0.12078044149610732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,256,0.035154667165544294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,6144,0.07352622350056966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,6144,0.10690310928556655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,16384,512,0.038225776619381376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,5120,0.06627021895514594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,5120,0.09141955773035686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,16384,0.23155644204881456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,4096,0.052296890152825244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,4096,0.07438488801320393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,65536,0.8832195599873861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,3584,0.049046221706602305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,3584,0.0668462249967787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,12288,0.17812711662716338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,3072,0.04068622324201796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,3072,0.059227552678849965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,10240,0.15494933393266466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,2560,0.034078223837746516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,2560,0.05070399906900194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,8192,0.12812088595496282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,2048,0.029275556405385334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,2048,0.04242222176657783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,7168,0.11868533823225234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,1536,0.023504889673656885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,1536,0.03566577699449327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,6144,0.10300444232092963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,1024,0.016948444975747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,1024,0.028799109988742407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,5120,0.09053333600362141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,768,0.014118222726715935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,768,0.024703111913469102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,4096,0.07756444480684069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,3584,0.07261777586407132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,512,0.01127822200457255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,512,0.022287110487620037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,256,0.008775110873911116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,256,0.018438221679793466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,3072,0.06595200300216675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,128,0.006609777609507243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,128,0.017200888858901132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,2560,0.06013688776228163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,64,0.0058266665372583605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,64,0.017476444443066914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,12288,32,0.006030222194062338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,12288,32,0.017793777916166518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,65536,0.6371475325690376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,65536,1.0079359478420682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,2048,0.05253244439760844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,16384,0.15804889467027453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,16384,0.2476595507727729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,1536,0.047760887278450855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,12288,0.1192293299569024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,12288,0.18511111206478545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,768,0.03862399856249491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,1024,0.041307555304633245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,10240,0.10208711359235977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,10240,0.1552657816145155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,8192,0.08717244201236302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,8192,0.1252639955944485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,512,0.03708088729116652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,7168,0.0743760002983941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,7168,0.11006400320265029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,256,0.0336791111363305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,6144,0.06543911165661283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,6144,0.09861955377790664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,12288,128,0.02972888946533203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,5120,0.055732442273033984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,5120,0.08227377467685275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,65536,1.2220097647772896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,4096,0.04515022370550367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,4096,0.06917688581678602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,16384,0.30884265899658203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,3584,0.04180888997183906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,3584,0.06250666909747653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,12288,0.23929243617587617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,3072,0.036363555325402155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,3072,0.0555973317888048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,8192,0.16989866892496744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,10240,0.20641422271728516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,2560,0.031698667340808444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,2560,0.046704888343811035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,2048,0.02628799941804674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,2048,0.04016622238688999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,7168,0.1528044409222073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,1536,0.021205332544114854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,1536,0.03355288836691115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,6144,0.1342702176835802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,1024,0.015340444114473132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,1024,0.027591110931502447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,5120,0.11377067036098903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,768,0.012835555606418185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,768,0.024311999479929607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,4096,0.09433866871727838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,512,0.010537777509954242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,3584,0.08786933289633857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,512,0.021871111459202234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,256,0.007874666816658443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,256,0.01831999917825063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,3072,0.07800177733103435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,128,0.006237333433495627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,128,0.01718933383623759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,2560,0.06309155623118083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,64,0.005495999836259418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,64,0.017469333277808297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,10240,32,0.005681777579916849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,10240,32,0.017613333132531907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,65536,0.46356799867418075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,2048,0.05587200323740641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,65536,0.8170853190951877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,16384,0.12460088729858398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,16384,0.19996178150177002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,1536,0.0486195551024543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,12288,0.0973786645465427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,12288,0.14868088563283285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,768,0.03926755653487311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,10240,0.0777182247903612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,10240,0.12394400437672932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,1024,0.04151733385192023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,8192,0.06422399812274508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,8192,0.10093955861197577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,256,0.031707555055618286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,7168,0.059755557113223605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,7168,0.0904008878601922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,128,0.028601778878106013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,6144,0.048841777775022716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,6144,0.07983733548058404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,10240,512,0.03707111212942336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,5120,0.04496711161401537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,5120,0.06732710864808825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,16384,0.1745928923288981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,4096,0.03615466753641764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,4096,0.05775288740793864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,65536,1.2054381900363498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,12288,0.1366515556971232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,3584,0.031806220610936485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,3584,0.0521360006597307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,3072,0.028528001573350694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,3072,0.04731733269161648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,10240,0.11899644798702663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,2560,0.02482044365670946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,2560,0.0396088891559177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,8192,0.10160711076524521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,2048,0.020138666033744812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,2048,0.03433155682351854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,7168,0.0934000015258789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,1536,0.016598222984208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,1536,0.02880000074704488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,6144,0.08421066734525894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,1024,0.012135999898115793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,1024,0.02366133365366194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,5120,0.07420800129572551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,768,0.010488000180986194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,768,0.021648888786633808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,3584,0.06257688999176025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,512,0.00850133349498113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,512,0.01942755613062117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,4096,0.06624355581071642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,256,0.006284444282452266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,2560,0.04965510964393616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,256,0.01664355562792884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,128,0.0051893335249688895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,128,0.01566755606068505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,3072,0.05545333358976576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,64,0.0047102222839991255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,64,0.015824889143308003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,8192,32,0.004928888960017098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,8192,32,0.016120000018013846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,65536,0.4400648805830214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,65536,0.7634799745347766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,2048,0.0427653325928582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,16384,0.11077155669530232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,16384,0.18851289484235975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,768,0.03502488798565335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,12288,0.08688533306121826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,12288,0.13950578371683756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,1536,0.04038044479158189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,10240,0.07520088884565565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,10240,0.11710133817460801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,1024,0.035750223530663386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,8192,0.05776888794369168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,8192,0.09465689129299587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,128,0.02753244505988227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,7168,0.05212533473968506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,7168,0.08471022049585979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,256,0.029626667499542236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,6144,0.04649066593911913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,6144,0.07577066951327853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,8192,512,0.033259554041756526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,5120,0.03820800118976169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,5120,0.06293244494332208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,16384,0.2169520060221354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,4096,0.03240088952912225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,4096,0.05452266666624281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,65536,0.8474231296115451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,3584,0.029394666353861492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,3584,0.04965510964393616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,12288,0.1657466623518202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,3072,0.02647200061215295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,3072,0.04446666770511203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,10240,0.1424880027770996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,2560,0.022801778382725183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,2560,0.03788622220357259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,8192,0.11780533525678848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,2048,0.01863911085658603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,2048,0.033344000577926636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,7168,0.10732799768447876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,1536,0.014688000082969666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,1536,0.027719111906157598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,6144,0.0935608877076043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,1024,0.011853333148691388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,1024,0.02331377731429206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,5120,0.08024444182713826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,768,0.010188444621033138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,768,0.020266667008399963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,3584,0.06554577747980754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,512,0.008466666771305932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,512,0.019110222657521565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,4096,0.07011022170384724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,256,0.00647555540005366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,256,0.0166275550921758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,2560,0.052766223748524986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,128,0.005613333235184352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,128,0.015652444627549913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,3072,0.05925333499908447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,64,0.005197333378924264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,64,0.015775110986497667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,7168,32,0.005422222117582957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,7168,32,0.01605688863330417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,65536,0.3527546723683675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,65536,0.7263635529412164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,2048,0.04649777876006233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,16384,0.10134400261773004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,16384,0.17794222301907012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,1536,0.04282933473587036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,12288,0.07814488808314006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,12288,0.13135111331939697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,768,0.03638044330808852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,10240,0.07430044147703382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,10240,0.10903822051154242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,512,0.03472977876663208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,1024,0.038057777616712786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,256,0.030305776331159804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,8192,0.04927911029921638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,8192,0.08816266722149319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,7168,0.04559466573927137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,7168,0.07973688840866089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,6144,0.04028266668319702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,6144,0.0712862213452657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,7168,128,0.028219554159376357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,5120,0.03403288788265652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,5120,0.059912886884477407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,65536,0.6287146674262153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,4096,0.027592889136738245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,4096,0.05145422286457486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,16384,0.16580177678002253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,3584,0.025182222326596577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,3584,0.047319110896852284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,12288,0.12995021873050266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,3072,0.02307022280163235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,3072,0.04225066635343763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,10240,0.11151111125946045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,2560,0.020833777056799993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,2560,0.03659288750754462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,8192,0.09440444575415717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,2048,0.01556711064444648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,2048,0.031089779403474595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,7168,0.08678666750590007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,1536,0.013159111142158508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,1536,0.02675466736157735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,6144,0.07796266343858507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,1024,0.010083555347389644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,1024,0.0225608895222346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,5120,0.06858933634228177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,768,0.008728000024954477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,768,0.01995733380317688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,4096,0.061498668458726674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,512,0.007297777467303806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,512,0.018191110756662156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,3584,0.05758221944173177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,256,0.005518222020732031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,256,0.016535111599498324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,3072,0.05279733406172859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,128,0.00464088883664873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,128,0.01551999979548984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,2560,0.048136889934539795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,64,0.004222222086456087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,64,0.015602666470739575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,6144,32,0.004436444491147995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,6144,32,0.01572711103492313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,65536,0.32469601101345485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,2048,0.042303999265034996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,768,0.034336888127856784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,1536,0.04015466570854187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,65536,0.698572423723009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,16384,0.08336977826224433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,16384,0.1702453295389811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,12288,0.06250311268700494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,12288,0.12731644842359754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,10240,0.054711110062069364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,1024,0.03579644362131754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,10240,0.10550933414035374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,8192,0.044031110074785024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,8192,0.08567021952735053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,512,0.03160088923242357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,7168,0.04048177931043837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,7168,0.07646933529112074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,256,0.029529776838090684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,6144,0.03491200009981791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,6144,0.06884088781144884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,6144,128,0.027610666222042505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,5120,0.029550221231248643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,5120,0.058260447449154325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,12288,0.15785866313510472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,16384,0.2058675554063585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,4096,0.025080889463424683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,65536,0.7874497837490506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,4096,0.05069777700636122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,3584,0.022720000810093347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,3584,0.04647555616166857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,3072,0.019741333193249173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,10240,0.13708444436391196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,3072,0.04085155659251743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,2560,0.0169840008020401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,8192,0.11228977309332953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,2560,0.035282668140199445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,2048,0.014249778456158109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,2048,0.030509332815806072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,7168,0.10240799850887722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,1536,0.01200711147652732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,1536,0.02641511129008399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,6144,0.09050044748518203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,1024,0.00924622184700436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,5120,0.07820266485214233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,1024,0.02163644466135237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,768,0.008026666939258575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,768,0.018957333432303537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,3584,0.06257155868742201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,512,0.00645955569214291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,512,0.017694221602545846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,4096,0.06702844301859538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,256,0.005055111315515307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,256,0.01626666718059116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,2560,0.048123555050955884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,3072,0.05615999963548449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,128,0.004378666480382283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,128,0.015408888459205627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,64,0.0041742221348815495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,64,0.015504888362354703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,5120,32,0.004276444514592488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,5120,32,0.015572445260153877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,65536,0.24540887938605416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,2048,0.04292622208595276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,65536,0.541961775885688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,16384,0.09961777925491333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,16384,0.13159999582502577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,1536,0.04003555576006571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,12288,0.07425333393944634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,12288,0.09906578063964844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,768,0.03273422188229031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,10240,0.0650142232577006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,10240,0.08316266536712646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,1024,0.03516088922818502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,8192,0.04928177926275465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,8192,0.0680275559425354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,512,0.031439112292395696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,7168,0.045160889625549316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,256,0.029391109943389893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,5120,128,0.02723377777470483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,7168,0.061252441671159535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,6144,0.039839999543295965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,6144,0.05606666538450453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,5120,0.03453066613939073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,5120,0.04808977908558316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,16384,0.15164000458187527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,4096,0.020367999871571858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,4096,0.04141777753829956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,65536,0.5602640046013726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,3584,0.018254222141371835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,3584,0.03745155533154806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,12288,0.11801244152916802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,3072,0.016937777400016785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,3072,0.03415022293726603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,10240,0.10311822096506755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,2560,0.014879110786649915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,2560,0.02976089053683811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,8192,0.0884666641553243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,2048,0.012142221960756512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,2048,0.026399999856948853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,7168,0.08141244120068021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,1536,0.010266666611035665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,1536,0.023365333676338196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,6144,0.0734764469994439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,1024,0.007770666645632849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,1024,0.019002666076024372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,5120,0.06541866726345487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,768,0.006870222174459034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,768,0.017439999514155917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,3584,0.05342488818698459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,512,0.005502222312821283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,512,0.01625066664483812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,4096,0.05755111243989733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,256,0.004495999879307217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,256,0.015110222829712762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,3072,0.04841777682304382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,128,0.003910222401221593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,128,0.014319111903508505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,2560,0.04399999976158142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,64,0.003583999971548716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,64,0.014208889669842191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,4096,32,0.003912000192536248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,4096,32,0.014292443792025248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,65536,0.22235822677612305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,65536,0.5224586592780219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,768,0.030332446098327637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,16384,0.06881688700781928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,16384,0.12560799386766222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,1536,0.037336889240476824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,12288,0.05405422382884555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,2048,0.03912533322970072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,12288,0.09373866849475437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,10240,0.043576889567905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,10240,0.07961066563924153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,1024,0.03338044550683763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,8192,0.03358666764365302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,8192,0.06422844198015001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,256,0.02836266822285122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,7168,0.030958222018347845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,7168,0.05887822310129801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,128,0.02612888813018799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,6144,0.028954668177498713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,6144,0.0535715553495619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,4096,512,0.029063112205929224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,5120,0.023261333505312603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,5120,0.04596177736918131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,65536,0.5440550910101997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,4096,0.019274666905403137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,16384,0.14819110764397517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,4096,0.03992799917856852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,3584,0.01696711116366916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,3584,0.03637689020898607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,12288,0.11645689275529648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,3072,0.015951111912727356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,3072,0.03296533226966858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,10240,0.10151910781860352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,2560,0.01420177850458357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,2560,0.028748446040683325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,8192,0.08719822433259751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,2048,0.0129457778400845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,2048,0.02609688871436649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,7168,0.08022222254011366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,1536,0.01125511113140318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,1536,0.022829333941141765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,6144,0.07250489128960504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,1024,0.008727110922336578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,1024,0.019174221489164565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,5120,0.06431377596325345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,768,0.0075128889746136135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,768,0.017668444249365065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,3584,0.052228444152408175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,512,0.005671111245950063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,512,0.016166221764352586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,2560,0.043655110730065234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,4096,0.05619733201132881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,256,0.00461422238085005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,256,0.01499555508295695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,128,0.004008888784382078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,128,0.014179555906189812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,3072,0.04843822121620178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,64,0.0037582222786214617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,64,0.0144213338692983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3584,32,0.003960000144110786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3584,32,0.014135110709402295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,65536,0.18673600090874565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,65536,0.48120445675320095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,2048,0.038951999611324735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,16384,0.059106667836507164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,16384,0.11780977249145508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,1536,0.037500444385740496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,768,0.030772444274690416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,12288,0.04295644495222303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,12288,0.08844711383183797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,10240,0.03862933317820231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,10240,0.07512711154090033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,1024,0.032425777779685125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,8192,0.029725333054860432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,8192,0.06202844116422865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,128,0.026157332791222468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,7168,0.028051555156707764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,7168,0.05632266733381483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,256,0.0280675556924608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,6144,0.023780445257822674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,6144,0.0513555539978875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3584,512,0.029102222787009344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,5120,0.02032888929049174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,5120,0.04372088776694404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,65536,0.7484711011250814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,4096,0.019119110372331407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,4096,0.038627554972966514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,16384,0.14268978436787924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,3584,0.014871110518773397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,3584,0.03457866774664985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,12288,0.11077955696317886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,3072,0.014296000202496847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,3072,0.03192266821861267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,10240,0.09558577669991387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,2560,0.011510222322411008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,2560,0.027609777119424608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,8192,0.08185599909888373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,2048,0.010999110837777456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,2048,0.02493688960870107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,7168,0.07538311349021064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,1536,0.008623999853928884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,1536,0.022121777137120564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,6144,0.06880799929300944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,1024,0.0069120000633928515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,1024,0.018530666828155518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,5120,0.061724444230397545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,768,0.0064533332155810455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,768,0.017258667283587985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,4096,0.054660442802641124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,512,0.005125333451562458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,512,0.016136000553766888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,3584,0.05133688780996534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,256,0.004080888711743885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,256,0.014852444330851236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,3072,0.0477804442246755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,128,0.003663111064169142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,128,0.014276444911956787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,2560,0.043439110120137535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,64,0.0035173334181308746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,64,0.014039110806253223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,3072,32,0.0036764442920684814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,3072,32,0.01408622165520986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,65536,0.16555199358198378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,65536,0.46084801355997723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,2048,0.03869600097338358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,16384,0.052667554881837636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,16384,0.11201777723100449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,1536,0.03695911169052124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,12288,0.039408889081743025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,12288,0.08515466584099664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,768,0.03018310997221205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,1024,0.031077331966824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,10240,0.033667554457982383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,10240,0.07271289163165622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,8192,0.026216889421145122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,8192,0.060234665870666504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,512,0.02943555514017741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,7168,0.024103111690945093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,7168,0.054888000090916954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,256,0.028466665082507666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,6144,0.021214221914609272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,6144,0.04991555545065138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,3072,128,0.026343110534879897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,5120,0.0184888889392217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,5120,0.04189244574970669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,65536,0.4948870870802138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,4096,0.015403555499182807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,4096,0.03569866551293267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,16384,0.13238933351304796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,3584,0.013895110951529609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,3584,0.033207999335394964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,12288,0.10334399673673843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,3072,0.011503111157152386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,3072,0.030591110388437908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,10240,0.09054399861229791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,2560,0.011047999891969891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,2560,0.026657778355810378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,8192,0.0778791109720866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,2048,0.008827555510732863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,2048,0.02420711186197069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,7168,0.07171199719111125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,1536,0.0075626663035816616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,1536,0.021684444612926904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,6144,0.06518133481343587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,1024,0.006060444646411472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,1024,0.018047110901938546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,5120,0.058563556936052114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,768,0.0052666668262746595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,768,0.01739022301303016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,4096,0.05043377810054355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,512,0.004576888763242298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,512,0.01603911154799991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,3584,0.048391110367245145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,256,0.003960888832807541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,256,0.014862222803963555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,3072,0.04533866710133023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,128,0.0035217776894569397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,128,0.014037332601017423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,2560,0.04135377870665656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,64,0.003399111123548614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,64,0.013975999421543546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,2048,0.036751998795403376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2560,32,0.003478222423129611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2560,32,0.013880000346236758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,65536,0.1296817726559109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,65536,0.4447840054829915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,16384,0.045239110787709556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,16384,0.10905422104729547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,1536,0.033435556623670794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,12288,0.048158221774631076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,1024,0.029969778325822618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,12288,0.0834862192471822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,10240,0.04045244389110141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,10240,0.07083200083838569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,768,0.028917332490285236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,8192,0.03193688723776076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,8192,0.05928622351752388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,512,0.028263999356163874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,7168,0.029631111356947158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,256,0.027585777971479628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,7168,0.05418311225043403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,6144,0.02546577817863888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2560,128,0.025576889514923096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,6144,0.048304888937208385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,5120,0.019352888067563374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,5120,0.040793776512145996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,65536,0.5024986796908909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,4096,0.014390221900410123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,16384,0.13662666744656032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,4096,0.034882667991850115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,3584,0.013014222184816996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,3584,0.032828443580203585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,12288,0.10599288675520156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,3072,0.011975111232863532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,3072,0.030180444320042927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,10240,0.09322666459613377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,2560,0.010462221999963125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,2560,0.026458667384253606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,8192,0.0797520014974806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,2048,0.00850666645500395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,2048,0.023815110325813293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,7168,0.07309866613811918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,1536,0.00740622232357661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,6144,0.06633333365122478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,1536,0.020530665914217632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,1024,0.005576000031497743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,1024,0.01798488861984677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,5120,0.05801600217819214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,768,0.004912000149488449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,768,0.0169804443915685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,3584,0.04799911048677233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,512,0.004242666479614046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,512,0.015909334023793537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,4096,0.0510844455824958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,256,0.003759110967318217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,256,0.014801777071423002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,3072,0.044453332821528115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,128,0.003350222276316749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,128,0.0140684445699056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,2560,0.04093066520161099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,64,0.0031795555518733132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,64,0.014044443766276041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,2048,32,0.0033359999457995095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,2048,32,0.013826666606797112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,65536,0.13298043939802381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,65536,0.4228462113274469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,2048,0.036159111393822566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,16384,0.03611200054486593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,16384,0.10602044396930271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,768,0.02861600120862325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,12288,0.03022222386466132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,12288,0.07968178060319689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,1536,0.03297955460018582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,10240,0.025832000705930922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,10240,0.0692524446381463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,1024,0.028935111231274072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,8192,0.020962667134073045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,8192,0.05684177743064033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,256,0.027448000179396734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,7168,0.018944000204404194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,7168,0.0521395570702023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,512,0.02807377775510152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,6144,0.016363556186358135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,6144,0.04549244377348158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,2048,128,0.025455110602908667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,5120,0.014386667145623101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,5120,0.03973599937227037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,65536,0.4822213384840224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,4096,0.011866666376590729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,4096,0.03532888823085361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,16384,0.1268799967235989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,3584,0.010871111518806882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,3584,0.032307555278142296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,12288,0.09798310862647162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,3072,0.009479111267460717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,3072,0.029790222644805908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,10240,0.08689866463343303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,2560,0.010109333528412713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,2560,0.026191110412279766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,8192,0.07456533114115398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,2048,0.008602666358153025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,2048,0.0225564440091451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,7168,0.06915555397669475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,1536,0.007126222054163615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,1536,0.01998488936159346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,6144,0.062021334966023765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,1024,0.005500444521506627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,5120,0.05490666627883911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,1024,0.017679111825095285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,768,0.00479288895924886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,768,0.016759999924235873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,4096,0.04824622140990364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,512,0.00416355538699362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,512,0.01573777695496877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,3584,0.047021332714292735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,256,0.0036355555057525635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,256,0.014878221684032016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,3072,0.04351644383536445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,128,0.0033475555893447665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,128,0.013969777358902825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,2560,0.03997688823276096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,64,0.003142222141226133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,64,0.01389422184891171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1536,32,0.0030586665703190696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1536,32,0.013636444177892474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,65536,0.1016648875342475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,2048,0.035154667165544294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,65536,0.4219004313151042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,16384,0.029245333539115057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,16384,0.10509955883026123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,1536,0.03277066681120131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,12288,0.02779910961786906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,12288,0.07980622185601129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,768,0.028441776831944782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,1024,0.029278222057554457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,10240,0.024578667349285547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,10240,0.06828888919618395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,8192,0.016492444607946608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,8192,0.056427558263142906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,512,0.02828977836502923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,7168,0.014902222487661572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,256,0.027408889598316614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,7168,0.051071110698911876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,6144,0.01328000012371275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,6144,0.045016888115141124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1536,128,0.025448888540267944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,5120,0.01181244436237547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,5120,0.038755555947621666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,65536,0.44855467478434247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,4096,0.010048888623714447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,4096,0.03378044565518697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,16384,0.11989333894517686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,3584,0.009324444664849175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,3584,0.031897776656680636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,12288,0.09313422441482544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,3072,0.00870222184393141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,3072,0.029376887612872656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,10240,0.08240266640981038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,2560,0.008404444489214156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,2560,0.024893333514531452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,8192,0.07060088713963826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,2048,0.0069217777086628815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,2048,0.022071111533376906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,7168,0.06574577755398221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,1536,0.005736888696750005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,1536,0.01979200045267741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,6144,0.05872799952824911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,1024,0.004599110947714912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,1024,0.01774044500456916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,5120,0.051910221576690674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,768,0.0041786668201287585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,768,0.016372443901167977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,4096,0.04571377899911669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,512,0.0037333332002162933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,512,0.015455111861228943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,3584,0.044922666417227856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,256,0.003391999958289994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,256,0.014415999253590902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,3072,0.04229244589805603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,128,0.0031102223114834893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,128,0.013733333183659447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,2560,0.036892443895339966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,64,0.0029608888758553397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,64,0.013742222554153867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,1024,32,0.003051555612021022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,1024,32,0.013607111242082385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,65536,0.06519822279612224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,2048,0.034522665871514216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,65536,0.4205173386467828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,16384,0.025027554896142747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,16384,0.10483466916614109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,1536,0.031846221950319074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,12288,0.020573332905769348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,12288,0.07829244269265069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,1024,0.02819733487235175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,10240,0.019369777705934312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,10240,0.0663084454006619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,768,0.0277839998404185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,8192,0.016184889607959323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,8192,0.05599200063281589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,512,0.027892443868849013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,7168,0.0147386673423979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,7168,0.05011644297175937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,256,0.027259555127885606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,6144,0.013184888495339287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,6144,0.04520533482233683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,1024,128,0.02535555594497257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,5120,0.011763555308183035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,5120,0.03879466652870178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,16384,0.11908088790045844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,4096,0.009847111172146266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,65536,0.4421688980526394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,4096,0.03357333276006911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,3584,0.009083555804358589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,3584,0.030437333716286555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,12288,0.09379644526375665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,3072,0.00849422232972251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,3072,0.028024888700909082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,10240,0.08198666572570801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,2560,0.007356444166766272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,2560,0.02437955637772878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,8192,0.07024266984727648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,2048,0.00629333323902554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,2048,0.021664889322386846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,7168,0.06472622023688422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,1536,0.005518222020732031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,1536,0.019389332996474374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,6144,0.05889955494138929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,1024,0.00462311092350218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,1024,0.017325333423084684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,5120,0.051710221502516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,768,0.004232888834344016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,768,0.016341333587964375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,4096,0.04529955652025011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,3584,0.04444800151718987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,512,0.0037120001183615792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,512,0.015276443627145557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,256,0.003302222324742211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,256,0.014305777019924588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,3072,0.04151644309361776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,128,0.003118222165438864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,128,0.013762666947311826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,2560,0.03705244594150119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,64,0.0029164445069101122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,64,0.013594666288958656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,768,32,0.002938666691382726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,768,32,0.01351466692156262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,65536,0.04922577738761902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,2048,0.032568001084857516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,65536,0.4205706649356418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,16384,0.017833777599864535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,16384,0.10549066464106242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,1536,0.031319111585617065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,12288,0.01610844499535031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,1024,0.028228445185555354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,12288,0.07819288969039917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,10240,0.014353776971499125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,10240,0.066685332192315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,768,0.027691556347741023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,8192,0.012828444441159567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,8192,0.05551822317971123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,512,0.027927110592524212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,7168,0.011855999628702799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,7168,0.05018488897217644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,256,0.026919111609458923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,6144,0.012099555797047086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,6144,0.04467111163669162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,768,128,0.025460445218616064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,5120,0.010785777535703448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,5120,0.0389395554860433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,65536,0.43234578768412274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,16384,0.11626933680640326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,4096,0.009026666482289633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,4096,0.033747557136747576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,3584,0.008239111138714684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,3584,0.03132177723778619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,12288,0.09172711107465957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,3072,0.007155555817815993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,10240,0.07983644141091241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,3072,0.02787644333309597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,2560,0.007381333245171442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,2560,0.02428177826934391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,8192,0.06901244322458903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,2048,0.006701333241330252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,2048,0.021948445174429152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,7168,0.06298400296105279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,1536,0.005525333185990651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,1536,0.019472888774342008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,6144,0.057370669311947294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,1024,0.004583111239804162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,1024,0.017200888858901132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,5120,0.051002668009863965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,768,0.0042088888585567474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,768,0.01644888851377699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,4096,0.04390666551060147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,3584,0.04412088791529337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,512,0.00370133337047365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,512,0.015299555328157214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,256,0.0032231110251612137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,256,0.014359111587206522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,3072,0.0410791105694241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,128,0.002942222274012036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,128,0.013677332964208392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,2560,0.03614666726854112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,64,0.0028444443725877335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,64,0.01366666704416275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,512,32,0.0028115556471877625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,512,32,0.013507555756303998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,65536,0.04494755466779073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,65536,0.42232799530029297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,2048,0.031686223215527005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,16384,0.011691555380821228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,16384,0.10578755537668864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,1536,0.03089155422316657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,12288,0.012196444802814059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,12288,0.07940977811813354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,1024,0.027657777070999146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,10240,0.011364444262451597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,10240,0.06735910971959432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,768,0.027683556079864502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,8192,0.009677333136399588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,8192,0.055552889903386436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,512,0.027273777458402846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,7168,0.00926577796538671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,7168,0.04931911163859897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,256,0.026743110683229234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,6144,0.00849333322710461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,6144,0.04416444566514757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,512,128,0.02511111067401038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,5120,0.00850933293501536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,5120,0.03804888990190294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,65536,0.42711912261115176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,4096,0.007639110916190677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,4096,0.03249511122703552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,16384,0.11526666747199164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,3584,0.007720000214046902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,3584,0.02923911147647434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,12288,0.08998666869269477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,3072,0.006890666567616993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,3072,0.027165333429972332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,10240,0.07882755332522921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,2560,0.006215111249023014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,8192,0.06806133190790813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,2560,0.024001777172088623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,2048,0.00543200017677413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,2048,0.02148977749877506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,7168,0.062021334966023765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,1536,0.00479644454187817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,1536,0.019327999817000497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,6144,0.05594933032989502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,1024,0.004060444318585926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,1024,0.017137777474191453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,5120,0.0500239994790819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,768,0.0037377778854635027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,768,0.016180444094869826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,4096,0.04359822140799629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,512,0.0034133332471052804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,512,0.01519555515713162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,3584,0.040878223048316106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,256,0.003100444459252887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,256,0.014130666851997375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,3072,0.03927910990185208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,128,0.002834666727317704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,128,0.013595555391576556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,2560,0.035270220703548856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,64,0.0027066665805048416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,64,0.013544888959990608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,2048,0.031805333164003156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,256,32,0.0027431110954946945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,256,32,0.01332800007528729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,65536,0.036072889963785805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,65536,0.4210302299923367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,16384,0.013117333253224691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,16384,0.10569244623184204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,1536,0.03057066599527995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,12288,0.010588444769382477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,12288,0.07874577575259738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,1024,0.0277173337009218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,10240,0.009507555928495195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,10240,0.06696799728605482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,768,0.027408889598316614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,8192,0.008364444805516137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,8192,0.05507377783457438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,512,0.027593778239356145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,7168,0.007962666451931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,7168,0.04926310976346334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,256,0.02683288852373759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,6144,0.007511110769377814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,6144,0.044396443499459155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,256,128,0.024846222665574815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,5120,0.007010666860474481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,5120,0.038047999143600464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,65536,0.42853334214952254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,4096,0.006656888872385025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,4096,0.03143466512362162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,16384,0.11452889442443848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,3584,0.0069884442620807225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,3584,0.028660443094041612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,12288,0.08980621894200642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,3072,0.006711999989218182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,10240,0.07909778091642591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,3072,0.026689777771631878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,2560,0.006167111297448476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,2560,0.023594665858480666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,8192,0.06748800145255195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,2048,0.005702222386995952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,2048,0.02125599980354309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,7168,0.06123377879460653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,1536,0.004790222065316306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,1536,0.019134221805466544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,6144,0.056490666336483426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,1024,0.004108444270160464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,1024,0.016920889417330425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,5120,0.04976800084114075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,768,0.0037191112836201987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,768,0.015954666667514376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,4096,0.04230666822857327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,512,0.0034373334298531213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,512,0.01516266663869222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,3584,0.04042844308747186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,256,0.003072000005178981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,256,0.013955555028385587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,3072,0.03779911001523336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,128,0.0028204443968004654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,128,0.013429332938459186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,2560,0.03550488750139872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,64,0.00272444449365139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,64,0.013341333303186627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,128,32,0.0027626665929953256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,128,32,0.013251555462678274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,2048,0.03141955534617106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,65536,0.029431111282772485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,16384,0.010718222293588849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,65536,0.4213591151767307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,16384,0.10557155476676093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,12288,0.008996444443861643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,12288,0.07864444123374091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,10240,0.008217777642938826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,1536,0.03046400017208523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,10240,0.06694933440950182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,8192,0.007511110769377814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,8192,0.05524533324771457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,7168,0.006892444358931647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,7168,0.049248887432946094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,6144,0.006478222293986215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,6144,0.04394311043951246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,5120,0.006895111252864202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,5120,0.03634666734271579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,4096,0.0063866666621632045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,4096,0.03109688891304864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,3584,0.00685511115524504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,3584,0.02891555428504944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,3072,0.006495111104514863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,3072,0.02683199942111969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,2560,0.005992888990375731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,2560,0.023777777949968975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,2048,0.0053191110491752625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,1024,0.027669333749347266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,2048,0.02127911150455475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,1536,0.004660444541109933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,1536,0.018975110517607797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,1024,0.004000888930426704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,768,0.027275555663638647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,1024,0.016903999778959487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,768,0.003608889049953885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,768,0.01572088897228241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,512,0.0033493333806594214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,512,0.014916444818178812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,256,0.0029484445436133277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,256,0.014056889547242058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,64,0.013360888593726687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,128,0.002793777734041214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,128,0.013411555853154926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,64,0.0026435556097163092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,64,32,0.002623999905255106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,64,32,0.013257777525318993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,65536,0.02715644405947791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,16384,0.008471999731328752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,65536,0.4240684509277344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,16384,0.1061368915769789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,12288,0.007622222105662028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,12288,0.07843289110395643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,10240,0.007081777685218387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,10240,0.06701333655251397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,8192,0.006801777829726537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,8192,0.054811557133992515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,7168,0.006500444478458828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,7168,0.049422221051322095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,6144,0.0063235556913746735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,6144,0.04362044400639004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,5120,0.006649777707126405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,3584,0.00675022229552269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,5120,0.03578844335344102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,4096,0.0063884444534778595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,512,0.027611555324660406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,4096,0.030869334936141968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,3584,0.028785778416527644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,3072,0.0064186664919058485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,3072,0.02665688925319248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,256,0.026548445224761963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,2560,0.006031999985376994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,2560,0.023902222514152527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,2048,0.005304000029961268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,2048,0.02125599980354309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,1536,0.004648888690604104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,1536,0.01897333396805657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,1024,0.00397777764333619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,1024,0.016891555653678048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,768,0.003705777641799715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,384,128,128,0.024829333027203877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,768,0.015747555428081088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,512,0.0033111110743549135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,512,0.014927999840842353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,256,0.0029813332690132987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,256,0.013905778527259827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,128,0.002773333340883255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,128,0.013632888595263163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,64,0.002609777781698439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,64,0.013225778109497495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,384,32,32,0.002598222138153182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,384,32,32,0.01314399970902337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,16384,0.45023197597927517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,16384,0.7532435523139106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,12288,0.5870150989956325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,12288,0.3524613380432129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,10240,0.4869253370496962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,10240,0.2976337803734673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,8192,0.3894933329688178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,8192,0.23953154351976183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,7168,0.3476746612124973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,7168,0.21511733531951904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,6144,0.34614133834838867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,6144,0.18865511152479383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,5120,0.2550853358374702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,5120,0.17648088932037354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,4096,0.20833067099253336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,4096,0.13347378042009142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,10240,0.3517706659105089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,12288,0.4174728923373752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,16384,0.514400906032986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,3584,0.18688266807132295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,3584,0.11602311664157444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,7168,0.26003821690877277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,8192,0.7384062343173556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,6144,0.23047823376125762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,5120,0.2004959980646769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,3072,0.16550578011406794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,3072,0.10245600011613633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,4096,0.16143377621968588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,2560,0.1403155591752794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,2048,0.07621244589487712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,2560,0.09025600221421982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,2048,0.12012710836198594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,1536,0.09130222267574734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,1536,0.06414577696058485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,1024,0.06431821982065837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,1024,0.05496889021661547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,768,0.05451911025577121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,768,0.04546666807598538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,512,0.040299554665883384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,512,0.03988800115055508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,3584,0.1431635618209839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,256,0.02926844358444214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,256,0.03201244605912103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,512,0.054076443115870156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,128,0.02208711041344537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,128,0.027496889233589172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,2560,0.11379110813140869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,64,0.016377778516875375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,64,0.029359999630186293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,65536,32,0.016733333468437195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,1536,0.08253777689403959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,65536,32,0.034098668230904475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,65536,0.7485155529446073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,65536,0.5778355598449707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,768,0.061279111438327365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,16384,0.19304977522956002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,3072,0.13013688723246256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,16384,0.15141510963439941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,12288,0.1507306628757053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,12288,0.11958666642506917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,10240,0.12522933218214247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,10240,0.10304711262385051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,2048,0.09705155425601536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,8192,0.10140622324413723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,8192,0.08459466695785522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,1024,0.06762399938371447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,7168,0.09074311123953925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,7168,0.07682400279574923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,256,0.04707022176848518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,6144,0.07914221949047513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,6144,0.0684337748421563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,65536,128,0.04014666544066535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,5120,0.06740622387992011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,5120,0.06061510907279121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,12288,0.15552177694108751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,16384,0.18830755021837023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,4096,0.05482933256361219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,4096,0.04858488837877909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,10240,0.12821243868933782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,3584,0.049736890527937144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,3584,0.045033779409196645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,65536,0.6917724609375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,3072,0.043846223089430064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,8192,0.10899022552702163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,3072,0.04137511054674784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,2560,0.037779556380377874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,2560,0.03710133168432448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,2048,0.031601776679356895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,2048,0.032255109813478254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,7168,0.09865244229634602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,1536,0.025586666332350835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,1536,0.027652444110976324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,6144,0.08883111344443427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,1024,0.017852443787786696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,1024,0.023152889476882085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,5120,0.08104711108737521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,768,0.01443911095460256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,768,0.0207004447778066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,3584,0.06366578075620863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,512,0.010874666273593903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,512,0.019535111056433786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,2560,0.05350666575961643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,256,0.008392888638708327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,256,0.016541333662139047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,4096,0.06798488563961454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,128,0.006214222146405115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,1536,0.043634666336907275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,128,0.015512888630231222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,3072,0.058380444844563804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,64,0.005286222116814719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,64,0.01577333278126187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,16384,32,0.005629333357016246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,16384,32,0.016352888610627916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,2048,0.049067556858062744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,65536,0.7945404582553439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,65536,0.5438533359103733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,16384,0.16585689120822483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,16384,0.14115377267201742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,12288,0.11123466491699219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,12288,0.11117866304185654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,768,0.03818044397566054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,10240,0.09380977683597141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,10240,0.09571199946933323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,1024,0.039108445247014366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,8192,0.07561066415574816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,8192,0.07857688930299547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,512,0.034556445148256086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,7168,0.07701333363850911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,7168,0.07134399811426799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,128,0.0292231109407213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,16384,256,0.03166133496496413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,6144,0.059261335266960986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,6144,0.06400622261895074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,5120,0.05055377880732218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,5120,0.05636088715659248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,16384,0.15904000070359972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,65536,0.5843475659688314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,4096,0.04185511006249321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,4096,0.04546311166551378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,3584,0.03783377673890855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,3584,0.04160088962978787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,12288,0.1277057727177938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,3072,0.033384889364242554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,3072,0.03700000047683716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,10240,0.11174489392174615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,2560,0.028872887293497723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,2560,0.03334310981962416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,8192,0.0950968861579895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,2048,0.02407199972205692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,2048,0.029537777105967205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,6144,0.08005955484178331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,7168,0.08655289146635269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,1536,0.01938488913906945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,1536,0.025461332665549383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,1024,0.013342222405804528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,5120,0.07233066691292657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,1024,0.022113778524928625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,768,0.011203555597199334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,768,0.020764445265134174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,3584,0.05755199988683065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,512,0.00924444446961085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,512,0.01960177719593048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,4096,0.060532444053226046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,256,0.006788444601827198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,256,0.016380444169044495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,3072,0.05237422386805216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,128,0.005125333451562458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,128,0.015255999233987598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,2560,0.04823644293679131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,64,0.004675555560323927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,64,0.015599111715952555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,768,0.035769777165518865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,12288,32,0.005050666630268097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,12288,32,0.015999111864301894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,65536,0.5239733590020074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,65536,0.5356968773735894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,16384,0.12446844577789307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,16384,0.13815289073520237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,1536,0.039846221605936684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,12288,0.09671111239327325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,12288,0.10898400015301174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,2048,0.0437084436416626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,10240,0.08021066586176555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,10240,0.09458311398824056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,1024,0.036574221319622464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,8192,0.06508800056245592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,8192,0.07685510979758368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,256,0.029818667305840388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,7168,0.05920888980229696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,7168,0.06948266426722209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,512,0.032885332902272545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,12288,128,0.02663555575741662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,6144,0.0518124434683058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,6144,0.06265511115392049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,5120,0.04408977760208977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,5120,0.055567999680837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,16384,0.1475786632961697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,4096,0.03673333260748122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,4096,0.044385777579413525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,65536,0.56804535124037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,12288,0.11936267217000325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,3584,0.033109333780076765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,3584,0.04051288962364197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,10240,0.10770844750934178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,3072,0.02952444553375244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,3072,0.036208887894948326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,2560,0.02564711040920681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,2560,0.032129777802361384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,8192,0.08971822261810303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,2048,0.021191999316215515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,2048,0.028508444627126057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,7168,0.08202311065461901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,1536,0.016919111212094624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,1536,0.024678222007221643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,6144,0.07633777459462483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,1024,0.012301333248615265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,1024,0.022052443689770166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,5120,0.06814044713973999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,768,0.010655110908879174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,768,0.020198222663667467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,3584,0.05494399865468343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,512,0.008810666700204214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,512,0.018192888961897958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,4096,0.057633777459462486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,256,0.006435555716355641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,256,0.01590755581855774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,2560,0.044999112685521446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,128,0.005066666752099991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,128,0.01516533394654592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,3072,0.05135822296142578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,64,0.004611555486917496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,64,0.015321777926550971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,10240,32,0.004847110973464118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,10240,32,0.01573422220018175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,65536,0.39846311675177676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,65536,0.38547645674811465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,2048,0.0418124430709415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,768,0.034993777672449745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,16384,0.09919022189246283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,16384,0.10104888677597046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,1536,0.037793778710895114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,12288,0.0760986672507392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,12288,0.08179199695587158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,10240,0.06595822175343831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,10240,0.07167644633187188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,1024,0.035248001416524254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,8192,0.0535662207338545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,8192,0.05939733319812351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,128,0.02645688917901781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,7168,0.047500444783104785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,256,0.028676443629794653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,7168,0.05542400148179796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,6144,0.042066666815016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,6144,0.050211555427975126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,10240,512,0.03206133180194431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,5120,0.03605066736539205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,5120,0.044128888183169894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,65536,0.47693602244059247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,4096,0.033188445700539485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,4096,0.03567911187807719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,16384,0.13384355439080134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,3584,0.02679288884003957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,3584,0.032610668076409235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,12288,0.10803555779986912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,3072,0.023749333288934495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,10240,0.09563910961151123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,3072,0.029598222838507757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,2560,0.020492444435755413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,2560,0.026830222871568467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,8192,0.08089155620998807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,2048,0.016857778032620747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,2048,0.02455377744303809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,7168,0.07462933328416613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,1536,0.01296622223324246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,1536,0.02158311174975501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,6144,0.06889333327611287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,1024,0.009862222605281407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,1024,0.019489778412712943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,5120,0.06226666768391927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,768,0.008453333543406593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,768,0.01848444508181678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,4096,0.05288355549176534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,512,0.007116444408893585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,512,0.01614755557643043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,3584,0.04930488930808174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,256,0.005333333379692501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,256,0.014880888991885714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,3072,0.045033779409196645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,2560,0.04196622305446201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,128,0.004435555388530095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,128,0.01424711114830441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,64,0.004039999925427967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,64,0.014170666535695394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,8192,32,0.004275555411974589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,8192,32,0.0144195556640625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,65536,0.3223031096988254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,65536,0.3712560070885553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,2048,0.038443555434544883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,16384,0.08645244439442952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,16384,0.09750666883256699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,1536,0.035651554663976036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,12288,0.06845155689451429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,12288,0.07834133174684313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,1024,0.03345955411593119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,10240,0.0588702228334215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,10240,0.06941777467727661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,768,0.033237334754731916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,512,0.031048887305789526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,8192,0.046647999021742076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,8192,0.05645510885450575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,256,0.028088887532552082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,7168,0.04175288809670342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,7168,0.05303200085957845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,6144,0.037463112009896174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,8192,128,0.025937777426507738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,6144,0.04787910977999369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,5120,0.03202755583657159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,5120,0.042394667863845825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,65536,0.4692942301432292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,4096,0.029725333054860432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,16384,0.12935733795166016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,4096,0.033987555238935686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,3584,0.023741333021057978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,3584,0.031298667192459106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,12288,0.1042319999800788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,3072,0.021299555897712708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,3072,0.028590222199757893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,10240,0.0913626684082879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,8192,0.07771288686328463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,2560,0.018437332577175565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,2560,0.025762667258580525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,7168,0.07165600193871392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,2048,0.015039111177126566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,2048,0.0232622226079305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,1536,0.011602666642930774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,6144,0.06616533464855619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,1536,0.021321778496106465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,1024,0.009015999734401703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,1024,0.019304000669055514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,5120,0.05996177593866984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,768,0.007885333564546373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,768,0.01752622259987725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,4096,0.050889776812659375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,3584,0.04798577891455757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,512,0.006737777756320105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,512,0.015847111741701763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,256,0.00489688871635331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,256,0.015049777097172208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,3072,0.04373777906099955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,128,0.0042160000238153665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,2560,0.04167644513977898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,128,0.01406311160988278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,2048,0.03874844312667847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,1536,0.03573244478967454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,64,0.0037866665257347953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,64,0.014137778017255994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,7168,32,0.004109333372778363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,7168,32,0.014173333843549093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,1024,0.03369155526161194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,65536,0.29483376608954537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,65536,0.3642248842451308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,768,0.033639109796947904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,16384,0.0751840008629693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,16384,0.09494400024414062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,12288,0.060814221700032554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,12288,0.07564710908465914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,10240,0.05717866950564914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,10240,0.06752889023886786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,8192,0.04144800040456984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,8192,0.055451555384529956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,512,0.029542220963372126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,256,0.02828977836502923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,7168,0.036306666003333196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,7168,0.05121066504054599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,6144,0.03374933203061422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,7168,128,0.026141333911154006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,6144,0.04647466540336609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,5120,0.02905155552758111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,5120,0.04109155469470554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,65536,0.4439982308281793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,4096,0.023342221975326538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,4096,0.03292266527811686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,16384,0.12886488437652588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,3584,0.021354667014545862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,3584,0.03054844339688619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,12288,0.11055022478103638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,3072,0.019103111492262948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,3072,0.02810133496920268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,10240,0.09267199701733059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,2560,0.016184889607959323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,2560,0.02541333271397485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,8192,0.07902755339940389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,2048,0.013033777475357056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,2048,0.0229751103454166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,7168,0.07325066460503472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,6144,0.06772444645563762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,1536,0.01051111105415556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,1536,0.021168000168270532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,1024,0.008193777667151557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,1024,0.019303111566437613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,5120,0.06136888927883572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,768,0.007251555720965068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,768,0.01668711172209846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,3584,0.04754399922158983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,512,0.006026666611433029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,512,0.015650666422314115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,4096,0.05143644412358602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,256,0.004583111239804162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,256,0.014799111419253878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,3072,0.04333333174387614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,2560,0.04090933336151971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,128,0.004022222426202563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,128,0.014146667387750415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,64,0.0036640001667870414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,64,0.013908444179428948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,6144,32,0.003919110943873723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,768,0.033017777734332614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,1536,0.035374220874574445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,65536,0.2727377679612902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,6144,32,0.01404088901148902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,65536,0.3572115633222792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,2048,0.03728977839152018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,16384,0.06653333372539945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,16384,0.09345510933134292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,12288,0.056235558456844754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,12288,0.07444444629881117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,1024,0.033209777540630765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,10240,0.05289511217011345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,10240,0.06461866696675618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,8192,0.03547022077772353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,8192,0.05407289001676771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,128,0.025467554728190105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,7168,0.03447733322779337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,256,0.027804444233576458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,7168,0.049441777997546725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,6144,0.030709332889980737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,6144,0.04310755597220527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,6144,512,0.028836445675955877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,5120,0.02732177740997738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,5120,0.03921511107020908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,65536,0.4223884476555719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,4096,0.02122933334774441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,16384,0.12329688337114121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,4096,0.03163999981350369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,3584,0.01925155520439148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,3584,0.02959022257063124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,12288,0.10099822282791138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,3072,0.017325333423084684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,3072,0.027472888429959614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,10240,0.08956178029378255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,2560,0.014793778459231058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,8192,0.0763982203271654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,2560,0.02497244377930959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,2048,0.011752000285519494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,2048,0.02271822260485755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,7168,0.07052710983488295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,1536,0.010128888818952773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,1536,0.02093066606256697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,6144,0.06488800048828125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,1024,0.007607999775144789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,1024,0.017901332841979135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,5120,0.05832622448603312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,768,0.006720888945791457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,768,0.016304888659053378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,3584,0.04572622312439812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,512,0.00554933316177792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,512,0.015864888827006023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,4096,0.04881244566705492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,256,0.004392888810899523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,256,0.014673777752452426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,2560,0.04029422336154514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,3072,0.04289688997798496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,128,0.0038337777886125776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,128,0.01400888959566752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,64,0.00360622215602133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,64,0.014083556003040738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,5120,32,0.003830222205983268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,5120,32,0.013976888524161445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,65536,0.21030488279130724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,2048,0.036483556032180786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,65536,0.350325345993042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,16384,0.062125331825680204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,16384,0.09037244319915771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,768,0.032327999671300255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,12288,0.05247999893294441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,1536,0.03498311175240411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,12288,0.0716257757610745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,10240,0.045167111688190036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,10240,0.06306222412321302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,1024,0.0325582226117452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,8192,0.03532177872127957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,8192,0.0518124434683058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,128,0.025337777203983728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,256,0.027443556321991816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,7168,0.03163199954562717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,7168,0.047008001142077975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,6144,0.02831288841035631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,6144,0.042265776130888194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,5120,512,0.027709333433045283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,5120,0.02269511090384589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,5120,0.037201778756247625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,65536,0.38731733957926434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,4096,0.01919555498494042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,16384,0.10864533318413629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,4096,0.031146665414174397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,3584,0.01684977776474423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,12288,0.089464889632331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,3584,0.028972443607118394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,3072,0.015623110863897534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,3072,0.0269813338915507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,10240,0.07899644639756945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,2560,0.012265777422322167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,2560,0.024519110719362896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,8192,0.06795199712117513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,2048,0.010808000134097205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,2048,0.022643556197484333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,7168,0.06334933307435778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,1536,0.008732444710201686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,1536,0.020505777663654752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,6144,0.05804889069663154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,5120,0.05271644393603007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,1024,0.007062222394678328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,1024,0.017464889420403373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,768,0.006007110906971826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,4096,0.04489511251449585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,3584,0.04285066657596164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,768,0.016591111818949383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,512,0.004801777915822135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,512,0.015749333633316886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,3072,0.039713776773876615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,256,0.004138666722509596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,256,0.014720888601409065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,128,0.003552000141806073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,2560,0.03770400087038676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,128,0.014056889547242058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,64,0.0033377777371141645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,64,0.0138586668504609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,4096,32,0.003527111063400904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,4096,32,0.013784888717863293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,65536,0.18391288651360405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,65536,0.33960355652703184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,2048,0.03502666619088914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,16384,0.05490666627883911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,16384,0.08724977572758992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,1536,0.03359555535846286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,12288,0.04407377706633674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,12288,0.06919733020994398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,1024,0.03164622187614441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,768,0.029320889049106177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,10240,0.03564622335963779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,10240,0.06153866979810926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,512,0.027085334062576294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,8192,0.028901331954532202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,256,0.027310222387313843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,8192,0.050682667228910655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,7168,0.02625422179698944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,7168,0.04619555672009786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,4096,128,0.025217778152889673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,6144,0.02344622214635213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,6144,0.041249778535630964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,5120,0.020256888535287645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,5120,0.03602844476699829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,65536,0.3913617663913303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,4096,0.017443555924627516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,16384,0.10865866475635105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,4096,0.030765334765116375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,3584,0.014704000618722705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,3584,0.02882933285501268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,12288,0.08798488643434312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,3072,0.013977777626779346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,3072,0.026755554808510676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,10240,0.07731378078460693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,2560,0.012064889073371887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,8192,0.06659911076227824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,2560,0.024587555064095393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,7168,0.06176622046364678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,2048,0.010451555252075195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,2048,0.02260977857642704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,6144,0.05718577570385403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,1536,0.008389333056079017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,1536,0.02033422225051456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,1024,0.006632888896597757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,5120,0.052128000391854175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,1024,0.01716088917520311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,768,0.00543200017677413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,768,0.016398222910033334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,4096,0.045003556542926364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,512,0.004393777913517422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,512,0.015399111641777886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,3584,0.04285066657596164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,256,0.003818666769398583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,256,0.014715555641386243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,3072,0.03961155480808682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,128,0.003403555601835251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,128,0.013961777091026306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,2560,0.037611554066340126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,64,0.0032248888164758682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,2048,0.035114665826161705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,64,0.013835555149449242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3584,32,0.003568888952334722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,1536,0.033906665113237165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3584,32,0.013744889034165276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,65536,0.15948977735307482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,65536,0.3386995527479384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,1024,0.03139200144343906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,16384,0.04912799927923414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,16384,0.0857342216703627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,768,0.031055112679799397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,12288,0.043271111117468945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,12288,0.0686640010939704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,10240,0.038219554556740656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,10240,0.06028888622919718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,8192,0.028074665202034846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,512,0.02772177755832672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,8192,0.04986933204862806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,7168,0.025437333517604407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,256,0.027475555737813313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,7168,0.045257776975631714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,6144,0.022424888279702928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,6144,0.040807998842663236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3584,128,0.025232000483406916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,5120,0.01943199998802609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,5120,0.035820444424947105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,65536,0.3867786725362142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,4096,0.015182221929232279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,16384,0.11179288228352864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,4096,0.03029866682158576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,12288,0.09199111329184638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,3584,0.013541333377361298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,3584,0.02862311071819729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,10240,0.0810391108194987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,3072,0.012538666526476542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,8192,0.06923377513885498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,3072,0.02618933386272854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,2560,0.010847999817795224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,2560,0.024168888727823894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,2048,0.009311999711725447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,2048,0.021959111094474792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,7168,0.06344088580873278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,1536,0.007799111306667328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,1536,0.018935110833909776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,6144,0.058252447181277804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,1024,0.006204444501135085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,1024,0.01698844465944502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,5120,0.05182844400405884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,768,0.00498844434817632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,768,0.016192000773217943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,4096,0.04435644547144572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,3584,0.04208088914553324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,512,0.0041848888827694785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,512,0.015418666932317944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,256,0.0036311112344264984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,3072,0.038875556654400296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,256,0.014545778433481852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,2560,0.03687644335958693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,128,0.0033191111352708605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,128,0.01387466655837165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,64,0.003119999956753519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,2048,0.03416444526778327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,64,0.013842666314707862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,3072,32,0.003335111050142182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,3072,32,0.013658666776286231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,1536,0.032630221711264715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,65536,0.138044450018141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,65536,0.33458132214016384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,768,0.02901511059867011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,1024,0.02909066610866123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,16384,0.0429057776927948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,16384,0.08364888694551255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,12288,0.03604088889227973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,12288,0.0665902230474684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,10240,0.03249511122703552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,10240,0.058634665277269155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,8192,0.022846221923828125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,8192,0.048410667313469775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,512,0.02661777701642778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,256,0.027007111244731482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,7168,0.020429333051045735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,3072,128,0.02486755616135067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,7168,0.043549332353803843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,6144,0.017599110802014668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,6144,0.03847288754251268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,5120,0.0177875558535258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,5120,0.03530577818552653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,65536,0.3617955578698053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,4096,0.013915555344687568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,16384,0.10144178072611491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,4096,0.030304001437293157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,12288,0.08202311065461901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,3584,0.01255555533700519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,10240,0.0727377798822191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,3584,0.028322666883468628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,3072,0.011520889070298938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,3072,0.026312889324294195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,2560,0.010019555687904358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,8192,0.06261244747373793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,2560,0.024308444725142583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,7168,0.058463997311062284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,2048,0.008748444418112436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,2048,0.021213332811991375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,6144,0.05322844452328152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,1536,0.007328888608349695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,1536,0.018608889646000333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,1024,0.005550222264395819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,5120,0.048692444960276283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,1024,0.016657777958446078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,4096,0.04277333286073473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,768,0.004591111093759537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,768,0.016027554869651794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,3584,0.04067111015319824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,512,0.004060444318585926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,512,0.015309333801269531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,256,0.003528000166018804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,3072,0.037821332613627114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,256,0.014449778530332776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,128,0.003216888962520493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,128,0.013629333012633853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,2560,0.035972442891862653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,64,0.003063999944263034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,64,0.013624889155228933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,2048,0.03378222386042277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2560,32,0.0032622222271230486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2560,32,0.013610666824711693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,65536,0.11124711566501194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,1536,0.030942221482594807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,65536,0.3330746756659614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,16384,0.03633688886960348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,16384,0.08251911401748657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,12288,0.03451199995146857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,12288,0.06555822160508898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,1024,0.02892355455292596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,10240,0.031176000833511353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,768,0.02903555499182807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,10240,0.05802488989300198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,512,0.026924444569481745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,8192,0.019288000133302476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,8192,0.04805688725577461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,7168,0.019536889261669584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,256,0.027223111854659185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,7168,0.04392888810899523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2560,128,0.025115556187099878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,6144,0.01811377704143524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,6144,0.03937155670589871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,5120,0.01609599921438429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,5120,0.03502666619088914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,65536,0.33595556683010525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,16384,0.09536266989178127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,4096,0.010804444551467896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,4096,0.029863999949561224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,12288,0.07825333542293973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,3584,0.009926222264766693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,3584,0.02778488728735182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,10240,0.0691137777434455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,3072,0.009202666580677032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,3072,0.026171555121739704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,2560,0.008094222181373173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,2560,0.023951111568344965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,8192,0.05957421991560194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,2048,0.007066667079925537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,2048,0.020766221814685397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,7168,0.05493866735034519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,1536,0.006248000181383557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,1536,0.018457776970333524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,6144,0.05056266652213203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,1024,0.004869333157936732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,1024,0.016728889611032274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,5120,0.04632355438338386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,768,0.004318222403526306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,768,0.01588088936275906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,4096,0.040786667002571955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,512,0.0038728887836138406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,3584,0.03900266687075297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,512,0.015207999282413058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,3072,0.03627289003796048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,256,0.003469333259595765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,256,0.014324444863531323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,128,0.003237333355678452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,2560,0.035411556561787925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,128,0.013566222455766467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,2048,0.032602667808532715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,64,0.003063111048605707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,1536,0.029823111163245306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,64,0.013638222383128272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,2048,32,0.0032053333189752367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,2048,32,0.013565333353148567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,65536,0.08657777971691555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,65536,0.33292444547017414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,1024,0.028237332900365193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,16384,0.03013422091801961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,16384,0.08265778091218737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,12288,0.026888888743188646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,12288,0.06532533301247491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,10240,0.023717333873112995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,768,0.028416001134448584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,10240,0.05733422438303629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,8192,0.016972444123691983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,8192,0.04737866587109036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,512,0.02624977793958452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,7168,0.015513777732849121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,7168,0.04285155402289497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,256,0.02625333269437154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,6144,0.013770666387346057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,2048,128,0.0244835548930698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,6144,0.03887910975350274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,5120,0.01426488823360867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,5120,0.03495999839570787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,65536,0.33185601234436035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,16384,0.09308266639709473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,4096,0.01074577785200543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,4096,0.030159999926884968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,12288,0.07605510950088501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,3584,0.00981866651111179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,3584,0.027894222074084814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,3072,0.008940444224410588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,10240,0.06787821981641981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,3072,0.02536444365978241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,2560,0.007992888490358988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,8192,0.05796800057093302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,2560,0.022813333405388728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,2048,0.00701333334048589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,2048,0.020280889338917203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,7168,0.05348799957169426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,1536,0.005799110978841782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,1536,0.018216888109842937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,6144,0.049448887507120766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,1024,0.004970666848950916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,1024,0.01649777756796943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,5120,0.04575644599066841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,768,0.004117333226733738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,768,0.015814221567577787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,4096,0.04045155644416809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,3584,0.038889778984917536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,512,0.003724444243643019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,512,0.015244444211324057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,256,0.003334222154484855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,256,0.014399111270904541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,3072,0.035642666949166193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,128,0.003031999907559819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,128,0.013526221944226159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,2560,0.033569776349597506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,64,0.0029004443850782183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,64,0.013528889252079858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1536,32,0.0030373332815037835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1536,32,0.013479111095269522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,65536,0.06537422206666735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,2048,0.031565331750445895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,65536,0.3324515554640028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,16384,0.02537333303027683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,16384,0.08280266655815972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,1536,0.029002666473388672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,12288,0.020119110743204754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,12288,0.06541688574684991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,1024,0.02786933382352193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,10240,0.018162666095627677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,10240,0.057237333721584745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,768,0.02790488799413045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,8192,0.013831111292044321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,8192,0.04702311091952854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,512,0.026429333620601233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,7168,0.012801777985360889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,7168,0.043057776159710355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,256,0.026629333694775898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,6144,0.014381332529915703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,6144,0.038607110579808555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1536,128,0.024757333927684363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,5120,0.013360888593726687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,5120,0.0347724457581838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,65536,0.31958044899834526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,4096,0.01087288889620039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,4096,0.028359111812379625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,16384,0.08941688802507188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,3584,0.009575111170609793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,3584,0.02630311085118188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,12288,0.07421955797407362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,3072,0.008988444175985124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,3072,0.024473778075642053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,10240,0.06581333610746595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,2560,0.007657777931955125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,2560,0.021893334057595994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,8192,0.0559555557039049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,2048,0.0064399999876817065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,2048,0.020045333438449435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,7168,0.0518817769156562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,1536,0.005568888866239124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,1536,0.018391110830836825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,6144,0.047931555244657725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,1024,0.0047217777205838096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,1024,0.01663733356528812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,5120,0.04503200120396084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,768,0.0041315555572509766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,768,0.01574844453069899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,4096,0.03968711031807794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,512,0.0036346668170558084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,512,0.01498222185505761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,3584,0.03764266769091288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,256,0.0031795555518733132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,256,0.014163555370436775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,3072,0.03447555502255758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,128,0.002945777856641345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,128,0.013447999954223633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,2560,0.031678222947650485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,64,0.0028719999310043124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,64,0.013487999637921652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,1024,32,0.0029395555870400537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,1024,32,0.01350044459104538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,65536,0.051510221428341336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,2048,0.03004177742534214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,65536,0.33201244142320424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,16384,0.018541332748201158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,16384,0.08301244179407756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,1536,0.028816888729731243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,12288,0.014988443917698331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,12288,0.06516177786721124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,1024,0.027783112393485174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,10240,0.014704000618722705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,10240,0.056110223134358726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,768,0.028024001253975764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,8192,0.012967999610635968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,8192,0.046457777420679726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,512,0.025920000341203477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,256,0.026230222649044458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,7168,0.01204711116022534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,7168,0.042230221960279674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,6144,0.01218933363755544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,6144,0.03832355472776625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,1024,128,0.024438222249348957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,5120,0.010684444672531553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,5120,0.03416088885731167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,65536,0.3179146713680691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,4096,0.009364444348547194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,16384,0.08973777956432766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,4096,0.02882755464977688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,3584,0.008488888541857401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,3584,0.02629866699377696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,12288,0.0741457806693183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,3072,0.007855110698276097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,3072,0.024316444993019104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,10240,0.06573600239223905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,2560,0.00701155596309238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,2560,0.0222044438123703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,8192,0.055567999680837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,2048,0.006319111006127463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,2048,0.02014311154683431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,7168,0.05089244577619764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,1536,0.00509688879052798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,1536,0.018170666363504197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,6144,0.0473671125041114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,1024,0.004348444441954295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,1024,0.016508445143699646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,5120,0.04432533184687296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,768,0.003969777789380816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,768,0.015673778123325773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,4096,0.03921511107020908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,512,0.003581333491537306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,3584,0.036221332020229764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,512,0.015030221806632148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,256,0.0032293332947625052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,256,0.014276444911956787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,3072,0.03410222133000692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,128,0.002961777771512667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,128,0.013658666776286231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,2560,0.0321359998650021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,64,0.0028320000403457214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,64,0.013608000344700284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,768,32,0.002867555452717675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,768,32,0.013677332964208392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,65536,0.04939466714859009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,2048,0.029558221499125164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,65536,0.33439289198981387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,16384,0.012975999878512489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,16384,0.08284444279140897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,1536,0.02851822310023838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,12288,0.01147911118136512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,12288,0.06503555509779188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,1024,0.027442667219373915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,10240,0.010519111322032081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,10240,0.05550933215353224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,768,0.027884443600972492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,8192,0.009383111364311641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,8192,0.046205331881841026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,512,0.026479111777411565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,7168,0.011589333415031433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,7168,0.04228444563017952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,256,0.026239110363854304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,6144,0.009408888717492422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,6144,0.037704888317320086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,768,128,0.024520888924598694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,5120,0.009783999787436591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,5120,0.03415022293726603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,65536,0.31532711452907985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,4096,0.008336000144481659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,4096,0.027995556592941284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,16384,0.08876888619528876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,3584,0.007849777738253275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,3584,0.026007999976476032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,12288,0.07203644514083862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,3072,0.006999111009968652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,3072,0.024344000551435683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,10240,0.06382844183180067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,2560,0.00628355559375551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,2560,0.02168888847033183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,8192,0.05475022064314949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,2048,0.005611555443869696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,2048,0.019863110449579027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,7168,0.05041689011785719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,1536,0.004883555488453971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,1536,0.01813155578242408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,6144,0.046457777420679726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,1024,0.004197333421972063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,1024,0.016338666280110676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,5120,0.044453332821528115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,768,0.0038542221817705366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,768,0.01564800077014499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,4096,0.03864444295565287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,512,0.003528000166018804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,3584,0.03637955586115519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,512,0.014993778533405729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,256,0.00313688897424274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,256,0.014225777652528552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,3072,0.033717334270477295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,2560,0.031097776359981958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,128,0.0029128889242808023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,128,0.013385777672131857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,64,0.0027822222974565295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,64,0.013462222284740873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,512,32,0.0027866665687825945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,2048,0.0295413335164388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,512,32,0.01331555595000585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,65536,0.03723999857902527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,65536,0.3313324451446533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,16384,0.011142222417725457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,16384,0.08255644639333089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,1536,0.02843911117977566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,12288,0.010103111465771994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,12288,0.06436089012357923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,1024,0.027170666389995154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,10240,0.00962311112218433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,10240,0.05520088805092705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,768,0.028007109959920246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,8192,0.008423111504978603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,8192,0.045922666788101196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,512,0.02583733366595374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,7168,0.00796088907453749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,7168,0.04133244355519613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,256,0.026338666677474976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,6144,0.009361777868535783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,6144,0.037399109866884016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,512,128,0.024289778537220422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,5120,0.008343111309740279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,5120,0.032821334070629544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,65536,0.313007116317749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,4096,0.007530666887760162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,4096,0.027473777532577515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,16384,0.08763200044631958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,3584,0.007068444457319047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,3584,0.025352888637118872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,12288,0.07217778099907769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,3072,0.00675022229552269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,3072,0.023955555425749883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,10240,0.06281689140531752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,2560,0.006191111273235745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,2560,0.02163822286658817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,8192,0.054274668296178184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,2048,0.005481777919663324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,2048,0.019973332683245342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,7168,0.049662222464879356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,1536,0.0047733332547876565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,1536,0.018177777528762817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,6144,0.044694221682018705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,1024,0.0040737779604064096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,1024,0.016415110892719693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,5120,0.0432604451974233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,768,0.003751111113362842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,768,0.015447111593352424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,4096,0.036304887798097395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,3584,0.03467911150720384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,512,0.0034142221427626083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,512,0.01479200025399526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,256,0.0031155554784668815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,256,0.014242667290899487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,3072,0.03278399838341607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,128,0.0028515555378463534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,128,0.013272888958454132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,2560,0.030986666679382324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,64,0.002765333279967308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,64,0.013255111045307584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,256,32,0.0029502221279674103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,256,32,0.013223999904261695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,2048,0.029283556673261855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,65536,0.025536888175540503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,65536,0.3358302116394043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,16384,0.010307555397351583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,16384,0.08303378025690715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,1536,0.028307557106018066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,12288,0.008852444589138031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,12288,0.06389422549141778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,1024,0.027451554934183758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,10240,0.007848888635635376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,10240,0.05496444304784139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,768,0.027593778239356145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,8192,0.007080888582600488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,8192,0.045478221442964345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,512,0.025614221890767414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,256,0.02612000041537815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,7168,0.006748444504208035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,7168,0.04065510961744521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,6144,0.006537777682145436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,6144,0.03559466534190708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,256,128,0.023963555693626404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,5120,0.006978666616810693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,5120,0.03196533189879523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,65536,0.3131715456644694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,4096,0.006539555473460092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,16384,0.08793599738015069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,4096,0.027439999911520217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,3584,0.0069777775141927935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,3584,0.025416000021828547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,12288,0.07137688663270739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,3072,0.006647111227114995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,10240,0.0632515549659729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,3072,0.0236151119073232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,2560,0.006011555592219035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,2560,0.021583999196688335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,8192,0.053801778290006846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,2048,0.005356444252861871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,2048,0.019738665885395475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,7168,0.04875733455022176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,1536,0.004604444321658877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,1536,0.017861333158281114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,6144,0.04460710949367947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,1024,0.003962666624122196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,1024,0.01607466737429301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,5120,0.040708445840411715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,768,0.0036515556275844574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,768,0.015391111373901367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,4096,0.03605422377586365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,512,0.003310222178697586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,512,0.014688889185587565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,3584,0.03449599941571554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,256,0.0029724445194005966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,256,0.013910222384664746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,3072,0.03219111098183526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,128,0.0028213332924577924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,128,0.013146667016877068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,2560,0.030993776188956365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,64,0.00263288886182838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,64,0.013145777914259167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,128,32,0.0026480000880029467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,128,32,0.01311644415060679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,2048,0.028914666838116113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,65536,0.021631111701329548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,16384,0.007296000089910295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,65536,0.33483465512593585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,12288,0.006709333509206772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,16384,0.08154133293363783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,12288,0.06386488676071167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,10240,0.006774222271309958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,10240,0.05486399928728739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,8192,0.006688000013430913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,8192,0.04428622126579285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,7168,0.006600889066855113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,7168,0.039847999811172485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,6144,0.0063999998900625445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,1536,0.028063999281989202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,6144,0.03537155522240533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,5120,0.006780444333950679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,5120,0.031656887796190046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,1024,0.02663555575741662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,4096,0.0063697778516345555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,4096,0.02755022214518653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,3584,0.006803555621041193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,3584,0.025232000483406916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,3072,0.006500444478458828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,3072,0.023455111516846552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,2560,0.0059262220230367445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,2560,0.021447999609841242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,2048,0.005299555758635203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,768,0.027079110344250996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,1024,0.003942222230964237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,2048,0.019742222295867074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,1024,0.016128000285890367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,1536,0.004580444345871608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,1536,0.0177831103404363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,768,0.0035866668654812705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,768,0.015354666444990369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,512,0.00328533330725299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,512,0.014575110541449653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,256,0.0029315555261241067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,256,0.013727111121018728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,128,0.0027359999302360746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,128,0.013213333156373767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,64,0.002570666579736604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,64,0.01313688854376475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,64,32,0.0026231110095977783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,64,32,0.013036444783210754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,65536,0.015917332635985482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,16384,0.006591111007663939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,512,0.02544622288809882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,65536,0.33304089970058864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,16384,0.08216533395979139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,8192,0.006568888823191325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,12288,0.006516444600290722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,12288,0.06386666827731662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,10240,0.00673333348499404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,10240,0.054568890068266124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,8192,0.044791999790403575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,7168,0.006383111079533895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,7168,0.04014844364590115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,256,0.025927111506462097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,6144,0.0062773335311147906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,6144,0.035192890299691096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,5120,0.006680888848172293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,5120,0.031697776582505964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,4096,0.006206222292449739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,2560,0.005998222364319696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,4096,0.027479110492600337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,3584,0.006709333509206772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,3584,0.02551555633544922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,3072,0.0063777777055899305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,3072,0.023553777072164748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,2560,0.021459556288189355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,2048,0.0053031109273433685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,1024,0.016120000018013846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,2048,0.019695111446910434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,1536,0.00462400002612008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,512,0.00328177772462368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,1024,0.00396799999806616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,1536,0.017954667409261067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,768,0.003645333151022593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,768,0.015255999233987598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,256,0.0029644444584846497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,512,0.014658666319317289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,128,0.0027831111931138565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,256,0.013811555173661975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,128,0.013240888714790344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,32,0.013085333009560903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,64,0.002604444407754474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,256,32,64,0.013078221844302284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,256,32,32,0.0026462222966882917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,16384,0.5471964412265354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,16384,0.7411253187391492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,12288,0.41199466917249894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,12288,0.5675973362392849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,10240,0.3571546607547336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,10240,0.4778204494052463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,8192,0.379822227689955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,8192,0.29238666428460014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,7168,0.33418843481275773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,7168,0.25600356525844997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,6144,0.3370195494757758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,6144,0.22339378462897408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,5120,0.24613422817654082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,5120,0.18650310569339326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,8192,0.3100444475809733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,10240,0.37804532051086426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,12288,0.4389386706882053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,4096,0.20145511627197266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,7168,0.28767289055718315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,4096,0.14846311675177679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,6144,0.24234843254089355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,3584,0.17878311210208467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,3584,0.13816355334387884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,16384,0.552712016635471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,3072,0.15675466590457493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,3072,0.11475555102030437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,5120,0.19722133212619355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,2560,0.13844533761342367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,2560,0.10018044710159302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,2048,0.11166755358378093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,4096,0.16462577713860407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,2048,0.08133688900205824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,3584,0.14336444271935359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,1536,0.08842666943868001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,1536,0.06425866815778944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,1024,0.0632737808757358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,1024,0.04865688747829861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,3072,0.14023378160264757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,768,0.0518488883972168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,2560,0.11763021681043838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,768,0.04229777720239428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,1024,0.06574044624964397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,2048,0.10035377740859985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,512,0.03984977801640829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,512,0.05297866794798109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,1536,0.08067111174265544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,512,0.03660888804329766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,768,0.06510755750868055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,256,0.024944000773959692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,256,0.02903999884923299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,128,0.018414222531848483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,256,0.04404533240530226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,128,0.026664000418451097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,64,0.014371555712487964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,65536,128,0.03753333290417989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,64,0.028402666250864666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,65536,32,0.014512888259357877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,65536,32,0.03317777646912469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,16384,0.18394933806525338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,16384,0.16352532969580755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,12288,0.1424728896882799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,65536,0.7108897633022733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,65536,0.6335831218295628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,12288,0.1288506719801161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,10240,0.11994489034016927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,10240,0.10899644427829319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,8192,0.0959928896692064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,8192,0.08967644638485378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,7168,0.08499822351667617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,12288,0.15011021826002333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,7168,0.08219733503129747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,16384,0.19018044736650255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,10240,0.13007999791039362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,6144,0.07532178031073676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,6144,0.07177866829766168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,8192,0.10940621958838569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,5120,0.06387733088599311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,5120,0.05999822086758084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,7168,0.1018844445546468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,4096,0.05281777845488655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,6144,0.08922221925523545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,3584,0.04673155479960971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,4096,0.05069955521159702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,5120,0.07678933276070489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,3584,0.04610844453175863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,4096,0.06513155831231011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,3072,0.041478223270840116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,3072,0.04150044586923387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,3584,0.061347557438744434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,2560,0.035735110441843666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,2560,0.03696799940533108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,3072,0.05730933613247342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,2048,0.02990666694111294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,2560,0.053038222922219165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,2048,0.03271733389960395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,1536,0.023960888385772705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,2048,0.047932446002960205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,1536,0.02776799930466546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,1024,0.016708445217874315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,1536,0.04312710960706075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,1024,0.02281422250800663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,768,0.013087111214796701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,1024,0.03914577762285868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,768,0.020959999826219346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,512,0.009879111415810054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,768,0.03673422336578369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,512,0.019634667370054457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,256,0.007381333245171442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,256,0.01679022279050615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,128,0.015260444747077094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,512,0.035049776236216225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,128,0.005642666584915585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,256,0.030962665875752766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,64,0.004807999978462855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,128,0.026719111535284255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,16384,32,0.005135111096832487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,64,0.015650666422314115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,16384,32,0.016550223032633465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,16384,65536,1.0618951585557725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,16384,0.14005155033535427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,65536,0.5714071061876085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,65536,0.5676968892415365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,16384,0.1448728905783759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,12288,0.11785244941711426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,12288,0.11291644308302139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,10240,0.09236622518963283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,16384,0.14814755651685926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,8192,0.07370222277111478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,10240,0.095487998591529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,12288,0.1255191167195638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,8192,0.07938488986757067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,7168,0.06555822160508898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,7168,0.07203288873036702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,10240,0.11555822690327962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,6144,0.06337244643105401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,8192,0.08949955304463704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,6144,0.065665774875217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,5120,0.05371110969119602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,7168,0.08238933483759563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,5120,0.050181332561704844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,4096,0.04024089044994778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,6144,0.07414311170578003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,4096,0.0459191103776296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,5120,0.06623466809590657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,3584,0.03618222143914964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,4096,0.05643022060394287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,3584,0.04195288817087809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,65536,0.5444106525844997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,3072,0.0321066677570343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,3072,0.03772888912094964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,3584,0.053816000620524086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,2560,0.027656000521447923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,3072,0.05130933390723335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,1536,0.018431110514534842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,2560,0.033732444047927856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,2048,0.022791999909612868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,1536,0.0392213331328498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,2560,0.048087110122044883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,2048,0.02999733222855462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,2048,0.042392889658610024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,1536,0.02528444429238637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,1024,0.012587555580668978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,1024,0.022181333767043218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,768,0.010424888796276517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,768,0.020764445265134174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,512,0.033568888902664185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,1024,0.03753688931465149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,512,0.008271111382378472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,768,0.03420266509056091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,512,0.019131556153297424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,256,0.006063111126422882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,256,0.01650311052799225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,128,0.004607111215591431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,64,0.015480889214409722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,256,0.029560890462663438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,128,0.015088889333936902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,64,0.004249777644872665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,12288,128,0.026015111141734656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,12288,32,0.004567999806669024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,12288,32,0.016026667422718473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,65536,0.4772791332668728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,16384,0.12472800413767497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,65536,0.5427341991000706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,16384,0.13738221592373318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,12288,0.1104560030831231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,12288,0.09683289130528767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,16384,0.13963288731045193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,10240,0.08787644571728176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,10240,0.09069777859581842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,12288,0.11234400007459854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,8192,0.06512177652782865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,8192,0.07562400235070123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,10240,0.09771288765801324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,7168,0.0621946652730306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,7168,0.06837599807315402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,8192,0.08411288923687404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,6144,0.055626663896772593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,5120,0.04566399918662178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,6144,0.06140266524420845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,7168,0.07820355229907565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,5120,0.0514284438557095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,4096,0.03530044356981913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,6144,0.07038044267230563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,5120,0.06251821915308635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,4096,0.04431644413206312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,65536,0.7700311342875162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,3072,0.03678133421474033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,3584,0.0337084432442983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,4096,0.05381244421005249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,3584,0.040319111612108015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,3072,0.02823022339079115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,3584,0.05157688922352261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,2560,0.025166221790843542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,3072,0.0502071115705702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,2560,0.03279022375742594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,2048,0.020247111717859905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,2560,0.04517599940299988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,2048,0.028639998700883653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,2048,0.040136890278922185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,1536,0.01622577839427524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,1536,0.024852444728215534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,768,0.0203448881705602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,1024,0.012331555287043253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,1536,0.03754399882422553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,1024,0.021271111236678228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,768,0.01017155581050449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,256,0.00566133318675889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,1024,0.035423109928766884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,512,0.007655111451943715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,256,0.02750933335887061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,768,0.033850666549470686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,128,0.01514311134815216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,128,0.025548444853888616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,512,0.01848088867134518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,10240,512,0.03242666522661845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,256,0.015982222225930955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,32,0.01567911108334859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,128,0.00452533322903845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,64,0.004097777936193678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,10240,64,0.01538044379817115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,65536,0.38706665568881565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,10240,32,0.00444533344772127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,65536,0.40146756172180176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,16384,0.0980328917503357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,16384,0.10240177975760566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,12288,0.07650488615036011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,12288,0.08003555403815375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,10240,0.06613510847091675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,10240,0.07249777846866183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,8192,0.0517119997077518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,8192,0.057271997133890785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,7168,0.046963555945290454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,7168,0.05230844351980421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,6144,0.0417208903365665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,6144,0.047783109876844615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,256,128,128,0.02416533397303687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,5120,0.035840001371171736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,5120,0.04059733284844293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,16384,0.12311466534932454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,12288,0.10078399711185032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,4096,0.027983999914593164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,4096,0.03657155566745334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,3584,0.02516888909869724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,3584,0.0325857765144772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,7168,0.07256622446907891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,8192,0.07807822359932794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,3072,0.02239111065864563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,3072,0.031183110343085393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,6144,0.06565688716040717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,2560,0.01958666741847992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,2560,0.027129777603679236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,10240,0.08976444270875718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,2048,0.015810666812790763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,65536,0.4324435657925076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,2048,0.02386133372783661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,1536,0.012127999630239276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,1536,0.02141777839925554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,1024,0.009318222602208456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,1024,0.01939644416173299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,5120,0.05836533175574409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,768,0.008036444584528605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,768,0.017872000734011333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,3584,0.04784533381462097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,512,0.006684444430801604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,512,0.016740444633695815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,4096,0.05057155423694187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,2560,0.04144533475240072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,256,0.00443111111720403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,256,0.014910222755538093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,1536,0.03554666704601712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,768,0.03239288926124573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,128,0.0038968887594011095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,128,0.013916444447305469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,64,0.003597333199448056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,64,0.013981334037250943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,8192,32,0.003936888857020272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,8192,32,0.014402666025691561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,3072,0.04475733306672838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,65536,0.34726667404174805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,65536,0.36087645424736875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,16384,0.10522133111953735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,16384,0.09631199969185723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,2048,0.03736889031198289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,12288,0.06613777743445502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,12288,0.07690933015611437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,16384,0.12856445047590467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,10240,0.0661297771665785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,10240,0.06596000327004327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,12288,0.10080889198515151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,8192,0.04502311017778185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,10240,0.08722933133443196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,8192,0.05522400140762329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,7168,0.04051822092798021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,65536,0.47546842363145614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,8192,0.07717333237330119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,1024,0.03401511245303684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,7168,0.050219555695851646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,6144,0.04692444536421034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,6144,0.03594222333696153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,7168,0.06850577725304498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,5120,0.03481600019666884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,4096,0.025247111916542053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,6144,0.0628053347269694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,4096,0.03386133246951633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,5120,0.03924355573124356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,5120,0.05553333295716179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,3584,0.022936888866954382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,4096,0.04845244354671902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,3584,0.03126666612095303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,3072,0.020568889048364427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,3584,0.04740533232688904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,3072,0.029161777761247423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,2560,0.017652443713612027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,2560,0.04079733292261759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,3072,0.043472889396879405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,2560,0.025575111309687298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,2048,0.014148443937301636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,2048,0.023373333944214716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,1536,0.011365333365069495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,2048,0.03715111149681939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,1536,0.021010667085647583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,1536,0.03530044356981913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,1024,0.00870577742656072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,1024,0.01902311046918233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,1024,0.03456177645259433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,512,0.006550222221348021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,768,0.007735110819339752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,768,0.017651556266678702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,768,0.031981332434548274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,512,0.01699911057949066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,256,0.004679999831649992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,512,0.029480000336964924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,256,0.014837332897716098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,128,0.004107555581463708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,256,0.02731288969516754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,128,0.014074666632546319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,7168,128,0.02527644402450985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,64,0.0037946667936113146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,64,0.014022222823566861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,7168,32,0.004084444294373195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,7168,32,0.014249778456158109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,65536,0.299216005537245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,65536,0.34883822335137266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,16384,0.09189422263039483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,16384,0.0837528904279073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,12288,0.05760711431503296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,12288,0.07300800085067749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,10240,0.05545066793759664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,10240,0.06253955761591594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,8192,0.03951999876234267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,8192,0.05232622226079305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,128,0.025115556187099878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,7168,0.03476622369554307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,7168,0.048600001467598804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,256,0.02684533264901903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,6144,0.03445155421892802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,6144,0.044087999396853976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,8192,512,0.031303998496797346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,5120,0.031215998861524794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,5120,0.03693777653906081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,16384,0.11273422506120469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,10240,0.07761066489749484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,12288,0.0880168875058492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,65536,0.3906071186065674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,4096,0.02223555578125848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,4096,0.03295733200179206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,3584,0.020097777247428894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,3584,0.029754665162828233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,8192,0.0675999985800849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,3072,0.019381332728597853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,3072,0.027428444888856676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,2560,0.01536622146765391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,2560,0.024697777297761705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,2048,0.012441777520709567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,2048,0.02275377843115065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,7168,0.06418133444256253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,1536,0.010226666927337646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,1536,0.02082577844460805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,6144,0.057895110713111035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,1024,0.008014221986134848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,1024,0.01867111192809211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,5120,0.05161422159936693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,768,0.0070053330726093715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,768,0.01761777698993683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,3584,0.04313777883847555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,4096,0.045296000109778516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,512,0.005687110953860813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,512,0.015980444020695157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,256,0.004122666600677702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,3072,0.041738665766186185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,2560,0.03944266504711575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,256,0.014736889137162102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,128,0.0037377778854635027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,128,0.01383733335468504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,2048,0.03613511059019301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,64,0.003415999934077263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,64,0.013966222604115805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,6144,32,0.003620444486538569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,6144,32,0.014003554979960123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,65536,0.24502934349907768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,1536,0.0351511107550727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,65536,0.33381067381964785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,16384,0.06489421923955281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,16384,0.08762844403584798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,1024,0.0339626669883728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,12288,0.05110488997565376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,12288,0.07006933291753133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,768,0.0317884451813168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,10240,0.04303377866744995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,10240,0.05930755535761515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,8192,0.03482222225930955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,8192,0.05079022381040785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,512,0.028837333122889202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,256,0.026743110683229234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,7168,0.034182220697402954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,6144,128,0.025058666865030926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,7168,0.045795556571748525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,6144,0.027573333846198186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,6144,0.04179288943608602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,5120,0.02383377816942003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,5120,0.03489066825972663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,16384,0.11981156137254502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,65536,0.5685440169440376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,12288,0.09477244483100043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,4096,0.01945955554644267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,4096,0.03112799922625224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,10240,0.0832444429397583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,3584,0.01790577835506863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,3584,0.028676443629794653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,8192,0.07133866680992974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,3072,0.015767110718621146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,3072,0.02644799980852339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,2560,0.013225778109497495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,2560,0.024076445235146418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,2048,0.010954666468832227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,2048,0.02212711175282796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,7168,0.06618933545218574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,1536,0.009212444225947062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,6144,0.06070311201943291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,1536,0.020398222737842135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,1024,0.007265778051482306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,5120,0.053242666853798755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,1024,0.017321777012613084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,768,0.0068400001360310455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,768,0.016711110870043438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,4096,0.04584355486763848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,512,0.005236444373925527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,512,0.01586844523747762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,3584,0.04264355699221293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,256,0.0041519999504089355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,256,0.014516444669829475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,3072,0.04173688756095039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,2560,0.03878844446606106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,128,0.003658666792843077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,128,0.014083556003040738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,64,0.00342755557762252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,2048,0.03559199968973795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,64,0.013953778478834363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,5120,32,0.0034497777620951333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,5120,32,0.01389422184891171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,65536,0.20003022087944877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,65536,0.3233342170715332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,16384,0.060108443101247154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,16384,0.084041780895657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,1536,0.033803555700514056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,12288,0.052456001440684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,12288,0.06715199682447645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,1024,0.03350044290224711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,768,0.03108711043993632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,10240,0.0509057773484124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,10240,0.05618133147557577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,8192,0.030883553955290053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,8192,0.048413332965638906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,512,0.02880000074704488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,7168,0.029876444074842665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,7168,0.04446844591034783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,256,0.02662044432428148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,6144,0.030449777841567993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,6144,0.03908266623814901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,5120,128,0.024958221448792353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,5120,0.021705778108702764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,5120,0.03367822368939718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,65536,0.3360942204793294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,4096,0.016689777374267578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,4096,0.02962133288383484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,16384,0.10715555482440525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,3584,0.01440622243616316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,3584,0.027858667903476294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,12288,0.08554399675793117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,3072,0.014012444350454541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,10240,0.07580888933605619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,3072,0.025468443830808003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,2560,0.01181244436237547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,2560,0.023642667465739783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,8192,0.06550933255089654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,2048,0.009840889109505547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,2048,0.021949332621362474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,7168,0.061434666315714516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,1536,0.008221333225568136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,1536,0.02014222244421641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,6144,0.05539466606246101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,1024,0.0063733334342638654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,1024,0.017401778035693698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,5120,0.04860711097717285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,768,0.0057546666098965546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,768,0.016505777835845947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,4096,0.042230221960279674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,512,0.004586666822433472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,512,0.015841777125994366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,3584,0.04056533177693685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,256,0.003863999827040566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,256,0.014581332604090372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,3072,0.038880000511805214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,128,0.0034257777863078644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,128,0.013898667362001208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,2560,0.03678044345643785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,64,0.0032035555276605818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,64,0.013796444568369122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,2048,0.03400889039039612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,4096,32,0.0034266666819651923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,4096,32,0.013637333280510373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,1536,0.03274577856063843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,65536,0.17566933896806505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,65536,0.3049226601918538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,768,0.028798222541809082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,16384,0.04825866553518507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,16384,0.08006311125225492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,1024,0.030413332912656996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,12288,0.037978665696250066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,12288,0.06493777698940702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,10240,0.03225244416130914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,10240,0.05511377917395698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,8192,0.03783999880154928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,8192,0.04673066735267639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,512,0.027727999620967444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,256,0.025972444150182936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,7168,0.023745778534147475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,7168,0.04176977939075894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,6144,0.029230223761664495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,4096,128,0.02458933326933119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,6144,0.03721066647105747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,5120,0.01794488893614875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,5120,0.03273066547181871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,65536,0.33198844061957467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,4096,0.015533333023389181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,4096,0.029354665014478896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,16384,0.10106311241785686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,12288,0.08408088816536798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,3584,0.01552177800072564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,3584,0.02697155541843838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,3072,0.015007999208238391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,3072,0.025591111845440332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,10240,0.07426044676038954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,2560,0.012912888493802814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,2560,0.023264888260099623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,8192,0.06412711408403185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,2048,0.010617777705192566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,2048,0.02130399975511763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,7168,0.061080886258019336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,1536,0.008858666651778752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,6144,0.05387111173735725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,1536,0.019751111666361492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,5120,0.047873778475655444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,1024,0.006437333507670297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,1024,0.017139555679427255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,768,0.005314666777849197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,768,0.016376889414257474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,3584,0.039932443035973444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,4096,0.04118311074044969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,512,0.004474666797452503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,512,0.015655999382336933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,256,0.0038044444388813446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,256,0.014473777678277759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,3072,0.038190222448772855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,128,0.0034968890249729156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,128,0.013735111388895245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,2560,0.03636799918280707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,64,0.0032906666811969546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,64,0.013705777625242868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3584,32,0.003419555516706573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3584,32,0.013645333548386892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,65536,0.152746664153205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,65536,0.30108621385362416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,2048,0.033696889877319336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,16384,0.04581244455443489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,16384,0.07811288701163398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,1536,0.0326257778538598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,12288,0.03352266550064087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,12288,0.06358844704098172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,768,0.02760266595416599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,10240,0.028400888045628864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,10240,0.05280888742870755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,1024,0.029666665527555678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,8192,0.03254933489693536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,8192,0.045679999722374805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,256,0.026026666164398193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,7168,0.02385155525472429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,7168,0.04084177811940511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,128,0.024329778220918443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,6144,0.02463111115826501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,6144,0.0370053317811754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3584,512,0.028000000450346205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,5120,0.01592177814907498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,5120,0.03236533204714457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,65536,0.4063546657562256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,16384,0.10865155855814616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,4096,0.015262222952312894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,4096,0.02864711152182685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,3584,0.011990221838156382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,3584,0.02700088918209076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,12288,0.0879439976480272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,3072,0.013999111122555204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,3072,0.025400888588693406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,10240,0.07724089092678495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,2560,0.011353777514563667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,2560,0.0232631117105484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,8192,0.06693600283728705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,2048,0.009351111120647853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,2048,0.021576889687114294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,7168,0.06238044632805718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,1536,0.008059555457697975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,1536,0.018926221463415358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,6144,0.05713688664966159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,1024,0.006226666685607698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,5120,0.050048887729644775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,1024,0.017098666893111337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,768,0.005103999955786599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,768,0.016344888342751395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,4096,0.0437466667758094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,512,0.004300444490379757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,3584,0.040369778871536255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,512,0.015576889117558798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,256,0.003802666647566689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,256,0.014554666148291694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,3072,0.03980088896221585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,128,0.0034577778230110803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,128,0.013897778259383308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,2560,0.037300444311565824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,64,0.0032044444233179092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,2048,0.034641779131359525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,64,0.013639999760521783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,3072,32,0.0033573332346147964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,3072,32,0.013590222431553734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,65536,0.13475467099083796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,65536,0.2883288860321045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,16384,0.042819556262758046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,16384,0.07538755734761556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,1536,0.032552000549104475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,12288,0.028917332490285236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,1024,0.030004445049497817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,12288,0.061590221193101674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,10240,0.02938133478164673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,768,0.028231110837724473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,10240,0.05238577723503113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,8192,0.020522667302025687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,8192,0.04496444596184624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,512,0.027890665663613215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,7168,0.020907556017239887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,256,0.026211554805437725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,7168,0.04043644335534837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,6144,0.01610311037964291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,6144,0.03661422265900506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,3072,128,0.02459555533197191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,5120,0.013617777989970313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,5120,0.032229334115982056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,65536,0.2779751088884142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,4096,0.013374221821626028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,16384,0.08495822217729355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,4096,0.02849600050184462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,3584,0.01200177768866221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,3584,0.026686223016844854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,10240,0.06222666634453667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,12288,0.07017511129379272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,3072,0.009472889204819998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,3072,0.025144888295067683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,8192,0.054528888728883534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,2560,0.009889778163697984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,2560,0.02312266661061181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,2048,0.007393777370452881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,2048,0.020640888147883944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,7168,0.0510062211089664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,1536,0.006400888992680444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,1536,0.018440888987647164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,6144,0.047276443905300565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,1024,0.004968889057636261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,1024,0.016750221451123554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,5120,0.04249511162439982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,768,0.004499555461936527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,768,0.016137777103318107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,4096,0.03754399882422553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,512,0.00398577791121271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,512,0.015228443675571017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,3584,0.03729688790109422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,256,0.003524444583389494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,256,0.014284445179833306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,3072,0.0358497765329149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,128,0.0032177778581778207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,128,0.013555555707878537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,2560,0.034521778424580894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,2048,0.03201955556869507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,64,0.0030293334275484085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,64,0.013617777989970313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2560,32,0.003077333379122946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2560,32,0.013551111022631327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,65536,0.11007999711566502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,65536,0.2892826663123237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,16384,0.04144800040456984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,16384,0.07427644729614258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,1536,0.030512889226277668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,12288,0.03215822246339586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,12288,0.06062755319807264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,1024,0.028444445795483057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,10240,0.028175112273957994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,10240,0.05186133252249824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,768,0.026904000176323786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,8192,0.019038221902317472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,8192,0.04412622253100077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,512,0.027288888891537983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,7168,0.016931555337376065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,256,0.025774222281244066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,7168,0.040277332067489624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,6144,0.014987554815080432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,6144,0.03638844357596503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2560,128,0.02403822210099962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,5120,0.017688888642523024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,5120,0.031862222486072116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,16384,0.08284444279140897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,65536,0.2719751199086507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,4096,0.014332445131407844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,4096,0.028021332290437486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,3584,0.012936000194814471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,3584,0.026571555270089045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,12288,0.06823466883765326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,3072,0.011572444604502784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,3072,0.0247715562582016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,10240,0.06040088997946846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,2560,0.010084444450007545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,2560,0.022638221581776936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,8192,0.053284443087048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,2048,0.008471999731328752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,2048,0.02001688877741496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,7168,0.050148444043265454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,1536,0.006521777974234686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,1536,0.018209778600268893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,6144,0.046021332343419395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,1024,0.0047004446387290955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,1024,0.0166631109184689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,5120,0.04211466511090597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,768,0.004193777839342753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,768,0.016032889485359192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,4096,0.037309332026375666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,512,0.0037911112109820047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,512,0.015298666225539314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,3584,0.03639555639690823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,256,0.0033484444850020935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,256,0.014193778236707052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,3072,0.035537779331207275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,128,0.003104888937539525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,128,0.013722666435771518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,2560,0.033992889854643084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,64,0.0029155556112527847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,64,0.013526221944226159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,2048,32,0.00309244439833694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,2048,32,0.013363555901580386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,65536,0.08938311205969916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,2048,0.0318248901102278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,65536,0.28852976693047416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,16384,0.029367112451129492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,16384,0.07420266336864896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,1536,0.029845333761639063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,12288,0.024080889092551336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,1024,0.028544889556037054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,12288,0.060362663533952504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,10240,0.021909332937664453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,10240,0.05130044288105435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,768,0.027307555079460144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,8192,0.018265777164035372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,8192,0.043615112702051796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,512,0.027431999643643696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,7168,0.016427555018001132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,7168,0.04031733340687222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,256,0.025685333543353613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,6144,0.013105777402718862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,6144,0.03640977872742547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,2048,128,0.024271999796231587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,5120,0.012906666431162091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,5120,0.031769778993394636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,65536,0.2699875566694471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,4096,0.011024889018800525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,16384,0.08000089062584771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,4096,0.028613332245084975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,3584,0.010110222631030613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,3584,0.026403556267420452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,12288,0.06654577785068087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,3072,0.009036444127559662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,3072,0.023941333095232647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,10240,0.05938666396670871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,2560,0.00811288919713762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,2560,0.021405332618289526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,8192,0.05155822303560045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,2048,0.007248889240953658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,2048,0.019833778341611225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,7168,0.049550222025977246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,1536,0.006272000157170826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,1536,0.018391110830836825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,6144,0.04520710971620348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,1024,0.004676444249020683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,1024,0.016671111186345417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,5120,0.04154755671819051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,768,0.004237333519591225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,768,0.015879111157523263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,4096,0.03686044282383389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,3584,0.03657599952485826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,512,0.0037120001183615792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,512,0.015244444211324057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,256,0.0033208889265855155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,256,0.014236445228258768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,3072,0.03541600041919284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,128,0.0030862221287356485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,128,0.013536888692114087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,2560,0.033738666110568576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,64,0.0028479999552170434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,64,0.01351999988158544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1536,32,0.0029128889242808023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1536,32,0.01332355539004008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,65536,0.06185511085722181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,65536,0.28673866060045033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,2048,0.0296382208665212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,16384,0.024688000480333965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,16384,0.07457600037256877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,1536,0.028880887561374243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,12288,0.019655999210145738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,12288,0.06035288837220934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,1024,0.028207110034094915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,10240,0.015963556038008798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,10240,0.05024977856212192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,768,0.026738666825824316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,8192,0.013682666752073499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,8192,0.04344444473584493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,512,0.027228444814682007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,7168,0.012686221963829465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,7168,0.039480888181262545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,256,0.025636444489161175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,6144,0.012113778127564324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,6144,0.03625333309173584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1536,128,0.02397777802414364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,5120,0.010619555910428366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,5120,0.031247112486097548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,65536,0.348291556040446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,4096,0.00923644420173433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,4096,0.02770311137040456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,16384,0.09722577863269383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,3584,0.008543999658690559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,3584,0.02554133368863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,12288,0.080677330493927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,3072,0.007596444752481248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,3072,0.02343822187847561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,10240,0.07001777489980061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,2560,0.007086222370465596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,2560,0.021347555849287245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,8192,0.0595360000928243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,2048,0.006511111226346757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,2048,0.019715555840068393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,7168,0.05622400177849663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,1536,0.005130666825506422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,1536,0.018100443813535903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,6144,0.05195466677347819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,1024,0.004325333154863781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,1024,0.016764443781640794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,5120,0.04572888877656725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,768,0.00397511116332478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,768,0.01590666671593984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,4096,0.04081333345837063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,512,0.0035253332720862497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,512,0.015150222513410779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,3584,0.03879200087653266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,256,0.0031991110493739447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,256,0.014139556222491793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,3072,0.03710577885309855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,128,0.002945777856641345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,128,0.013617777989970313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,2560,0.034563554657830134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,64,0.0028239999794297745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,64,0.013451555536852943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,1024,32,0.0028355556229750314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,1024,32,0.01329955541425281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,65536,0.05119199885262383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,65536,0.28776176770528156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,2048,0.030856887499491375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,16384,0.017451556192504037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,16384,0.07393866777420044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,1536,0.029836446046829224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,12288,0.01718222267097897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,12288,0.05896266963746813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,1024,0.029118223322762385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,10240,0.015372445185979208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,10240,0.05012177758746677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,768,0.027072888281610277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,8192,0.013345777988433838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,8192,0.042585776911841504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,512,0.027503111296229895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,7168,0.01349688900841607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,7168,0.0396364430586497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,256,0.0255022214518653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,6144,0.012242666549152799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,6144,0.03628799981541104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,1024,128,0.024047111471494038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,5120,0.010830221904648675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,5120,0.031203554736243352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,65536,0.3190924326578776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,4096,0.00923822240697013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,4096,0.0266311119000117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,16384,0.09465066591898601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,3584,0.008586666650242275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,3584,0.025124443901909724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,12288,0.07711288664076063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,3072,0.007878222399287755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,10240,0.06663999954859416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,3072,0.023160000642140705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,2560,0.006713777780532837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,2560,0.02130488885773553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,8192,0.05800355805291069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,2048,0.005678222411208683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,2048,0.01979733341270023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,7168,0.05498933460977343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,1536,0.005059555586841372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,1536,0.0181697772608863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,6144,0.050923556089401245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,1024,0.004413333204057482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,1024,0.016795555750528973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,5120,0.04515999886724684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,768,0.004043555508057277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,768,0.015998222761683993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,4096,0.040251556369993426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,512,0.0036346668170558084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,512,0.01516711049609714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,3584,0.03605422377586365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,256,0.0032097777972618737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,256,0.01421688828203413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,3072,0.03548266821437412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,128,0.0029377777957253983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,2560,0.03292799989382426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,128,0.013566222455766467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,64,0.0027840000887711844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,64,0.013564444250530668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,768,32,0.0027439999911520216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,768,32,0.013377777404255338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,65536,0.038791110118230186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,65536,0.288611544503106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,2048,0.030382222599453394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,16384,0.014127110441525778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,16384,0.07371555434332953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,1536,0.029586666160159644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,12288,0.01220444424284829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,12288,0.05885244740380181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,1024,0.028519110547171697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,10240,0.01221333361334271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,10240,0.05009599857860141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,768,0.027124444643656414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,8192,0.010844444235165914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,8192,0.04298311140802172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,512,0.02720444401105245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,7168,0.01035911093155543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,7168,0.03936800029542711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,256,0.025729777084456548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,6144,0.00959022260374493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,6144,0.03558222121662564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,768,128,0.023915555742051866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,5120,0.009289777941173976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,5120,0.03009866674741109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,65536,0.2548000017801921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,4096,0.007860444486141205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,4096,0.02668711046377818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,16384,0.07592088646358915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,3584,0.007658667034573025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,3584,0.024466666910383437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,12288,0.0624835557407803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,3072,0.006843555718660355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,3072,0.02296177711751726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,10240,0.0549777779314253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,2560,0.006215999937719769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,2560,0.021167111065652635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,8192,0.04814844330151876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,2048,0.005478222337034013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,2048,0.019547555181715224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,7168,0.04681422313054403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,1536,0.0048142220411035745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,1536,0.018028444714016385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,6144,0.04333510994911194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,1024,0.004106666478845808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,1024,0.016419554750124615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,5120,0.0394755568769243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,768,0.0037564444873068067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,768,0.015763555963834126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,4096,0.03458399905098809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,512,0.0034497777620951333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,512,0.0151573336786694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,3584,0.034274667501449585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,256,0.003061333257291052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,256,0.01402311192618476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,3072,0.03200888964864943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,128,0.002830222249031067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,128,0.013593778014183044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,2560,0.030921777089436848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,64,0.0027217778066794076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,64,0.013305777476893531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,512,32,0.00272444449365139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,512,32,0.013228444589508904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,65536,0.034472889370388456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,65536,0.289955563015408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,2048,0.028492444091373022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,16384,0.013825777504179211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,16384,0.07330666648017035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,1536,0.028218666712443035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,12288,0.011572444604502784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,12288,0.05825155311160617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,1024,0.02773777809407976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,10240,0.010748444332016839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,10240,0.04965244399176704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,768,0.026464000344276428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,8192,0.009509333305888707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,8192,0.04202222161822849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,512,0.026914666096369427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,7168,0.00889155517021815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,7168,0.038937777280807495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,256,0.02534488836924235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,6144,0.008205333517657386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,6144,0.03508177730772231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,512,128,0.02383111086156633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,5120,0.0073928890956772705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,5120,0.029789331886503432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,65536,0.2543697886996799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,4096,0.006792000184456508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,4096,0.025943999489148457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,16384,0.07488355371687147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,3584,0.006994666324721442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,3584,0.024131556351979572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,12288,0.06186578008863661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,3072,0.006741333338949416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,3072,0.022636445032225713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,10240,0.05458133419354757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,2560,0.00619555554456181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,2560,0.020818667279349435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,8192,0.04703022374047173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,2048,0.005489777773618698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,2048,0.0191920002301534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,7168,0.045963555574417114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,1536,0.00479911102188958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,1536,0.01772177716096242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,6144,0.042373332712385386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,5120,0.03776977790726556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,1024,0.004041777716742622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,1024,0.01607466737429301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,768,0.0036862223512596558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,768,0.015434667468070984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,4096,0.03341689043574863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,512,0.0034231110993358824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,512,0.014630221658282809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,3584,0.03241511185963949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,256,0.003044444446762403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,256,0.01384888920519087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,3072,0.03199022346072727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,2560,0.030838221311569214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,128,0.0028560000161329904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,128,0.013325333595275879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,64,0.002704888789190186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,64,0.01330133361948861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,256,32,0.0026986667265494666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,256,32,0.013110222087966071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,65536,0.02548533346917894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,65536,0.28973423110114205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,2048,0.028431998358832464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,16384,0.009405333134863112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,16384,0.07272266679339938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,1536,0.027967111931906805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,12288,0.00812266684240765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,12288,0.05749155415429009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,1024,0.027842667367723253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,768,0.026458667384253606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,10240,0.0074275558193524676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,10240,0.04810933272043864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,8192,0.0069395556218094295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,8192,0.04080266753832499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,512,0.026688888669013977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,7168,0.006816888848940532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,7168,0.0373271107673645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,256,0.02518400053183238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,6144,0.006506666541099548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,6144,0.03371466530693902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,256,128,0.02379377848572201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,5120,0.0069422221018208405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,5120,0.02916711237695482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,65536,0.2537288930681017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,4096,0.006443555570311016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,4096,0.025593777497609455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,16384,0.07490577962663439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,3584,0.0069520001610120135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,3584,0.024002666274706524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,12288,0.061767107910580106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,3072,0.006717333363162146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,3072,0.022580444812774658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,10240,0.0540960000620948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,2560,0.005949333310127258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,2560,0.020583111378881667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,8192,0.0467564430501726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,2048,0.005286222116814719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,2048,0.018955555227067735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,7168,0.045624001158608325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,1536,0.004679111142953237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,1536,0.017479111750920612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,6144,0.04197600152757433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,1024,0.00397511116332478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,1024,0.015979554918077257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,5120,0.036963555547926165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,768,0.0036391110883818734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,768,0.015184000134468079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,4096,0.03292266527811686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,512,0.003302222324742211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,512,0.014451555079884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,3584,0.03231466809908549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,256,0.0029608888758553397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,256,0.013759999639458127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,3072,0.031850665807724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,128,0.002781333401799202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,128,0.013247999880048962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,2560,0.030396444929970637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,64,0.002604444407754474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,64,0.013170666992664337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,128,32,0.0026026666164398193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,2048,0.0284586681260003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,128,32,0.012878222597969903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,65536,0.01425244410832723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,16384,0.007803555164072249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,65536,0.28841866387261283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,16384,0.0728257762061225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,12288,0.0069164443347189165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,12288,0.05719466341866387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,10240,0.006805333412355847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,10240,0.04819644490877787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,8192,0.006641777853171031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,7168,0.006549333532651265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,8192,0.039855112632115684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,6144,0.006333333336644703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,7168,0.03639555639690823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,6144,0.033210668298933245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,5120,0.006743999818960826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,4096,0.006321777900060018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,5120,0.029097778929604426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,4096,0.02558311157756382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,3584,0.006747555401590135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,3072,0.006426666759782367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,3584,0.02390044430891673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,1536,0.027951111396153767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,3072,0.02221511138810052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,2560,0.005902222047249476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,2560,0.020795555578337777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,2048,0.005254222287072076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,1536,0.004629333400064045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,2048,0.018968888454967074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,1536,0.017258667283587985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,1024,0.003951999876234266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,768,0.0035706667436493766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,1024,0.015820443630218506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,768,0.015132443772421943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,512,0.0032675556010670135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,512,0.014411555396185981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,256,0.002926222152180142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,256,0.01368622233470281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,128,0.0027431110954946945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,128,0.013063111239009433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,64,0.0026177778426143858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,64,0.01297155519326528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,64,32,0.0025946667624844448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,64,32,0.012920888761679331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,65536,0.013170666992664337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,16384,0.006711111300521427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,65536,0.2885591189066569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,16384,0.07240088780721028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,12288,0.006484444356626934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,12288,0.05621600151062012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,10240,0.006745777610275481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,10240,0.04718666606479221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,8192,0.006624889042642381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,8192,0.03974489039844937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,7168,0.006480888773997624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,1024,0.02770400047302246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,6144,0.006252444452709622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,7168,0.03672444489267137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,6144,0.033230221933788724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,5120,0.006701333241330252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,5120,0.029112001260121662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,3584,0.006754666566848755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,4096,0.006251555350091722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,3072,0.006411555740568373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,4096,0.02558044426971012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,2560,0.005966222120655908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,3584,0.023903111616770428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,3072,0.022403554783927068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,2560,0.02049777739577823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,2048,0.0052986666560173035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,768,0.02606755495071411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,2048,0.018958222534921434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,1536,0.004631999880075455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,1536,0.017326222525702584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,1024,0.003958222352796131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,1024,0.015847111741701763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,768,0.003559999995761447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,768,0.015257777439223396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,512,0.0032400000426504347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,512,0.014528888795110913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,256,0.0029271110478374693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,256,0.013733333183659447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,128,0.002744888886809349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,32,0.002568888788421949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,128,0.013105777402718862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,192,32,64,0.0026062221990691293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,512,0.026521777113278706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,64,0.013040000365840064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,192,32,32,0.01295466638273663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,16384,0.7387137942843967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,16384,0.5947279930114746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,12288,0.5643333329094781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,12288,0.4510400030348036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,10240,0.4726044336954753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,10240,0.3944684399498834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,8192,0.3807066546546088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,8192,0.3234479957156711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,256,0.025099555651346844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,7168,0.36674756473965114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,7168,0.2887173228793674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,6144,0.2881688806745741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,6144,0.24543910556369355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,5120,0.2432373364766439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,5120,0.20713155799441865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,8192,0.3473840024736192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,10240,0.4180142349667019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,192,128,128,0.023764444722069636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,12288,0.4775893423292372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,6144,0.25142399470011395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,16384,0.6055493354797363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,7168,0.3112204339769152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,4096,0.19777688715193006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,4096,0.1677208873960707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,3072,0.15086400508880615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,3584,0.17329511377546522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,3584,0.1494657728407118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,2560,0.1391999986436632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,3072,0.12853333685133192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,3584,0.16389155387878418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,4096,0.1826444466908773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,5120,0.2565048800574409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,2560,0.11047910981708103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,2048,0.10468266407648723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,1536,0.0706346697277493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,2560,0.1259724431567722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,2048,0.0906497769885593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,3072,0.1443342235353258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,1536,0.08217510912153456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,1024,0.05842133363087972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,2048,0.10585155751970078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,1024,0.047760887278450855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,1536,0.08587821986940171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,768,0.05880977710088094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,768,0.04639200038380093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,1024,0.06626400020387438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,768,0.04280266828007168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,512,0.03495288888613383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,512,0.035767998960283064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,256,0.021506667137145996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,512,0.05103466577000088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,256,0.028869334194395278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,128,0.013398222625255585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,256,0.042416890462239586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,128,0.026564444104830425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,64,0.010320888625250923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,65536,128,0.036278221342298723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,64,0.028043554888831243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,65536,32,0.009931555224789513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,65536,32,0.03198933270242479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,16384,0.18168621593051484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,16384,0.17245154910617402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,65536,0.7584026654561361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,65536,0.64883820215861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,12288,0.13965600066714817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,12288,0.13402311007181802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,10240,0.13719911045498318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,10240,0.11936533451080322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,8192,0.09426399734285142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,8192,0.09662399689356486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,7168,0.08343288633558486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,16384,0.19883288277520073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,12288,0.1592586702770657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,7168,0.08701155583063762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,6144,0.07371111048592462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,5120,0.06219644678963555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,10240,0.13962488704257542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,6144,0.07554666863547431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,8192,0.11363111601935492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,7168,0.10419466760423447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,5120,0.06392977635065715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,6144,0.09078222513198853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,4096,0.05094666613472832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,4096,0.05405333307054308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,5120,0.0798542234632704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,3584,0.045781334241231285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,3584,0.0496782230006324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,4096,0.06797511047787137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,3072,0.04034666551484002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,3584,0.06376977761586507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,3072,0.04366577665011088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,2560,0.034875555170906916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,2048,0.034529778692457415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,3072,0.05974577532874214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,2560,0.03912533322970072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,2048,0.028832889265484277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,2560,0.0549777779314253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,1536,0.023336889015303716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,65536,0.6996764606899686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,2048,0.04907555712593926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,1536,0.02983377708329095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,1024,0.016349333855840895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,1536,0.04600444436073303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,1024,0.02329155637158288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,768,0.01290933373901579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,1024,0.03903733359442817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,768,0.021032000581423443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,768,0.0370435549153222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,512,0.00962577760219574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,512,0.019519110520680744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,256,0.0070293330483966405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,512,0.03407377666897244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,256,0.01685333251953125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,32,0.005030222237110138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,128,0.005098666581842634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,256,0.030279109875361126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,128,0.015288000305493673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,16384,64,0.00453599997692638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,16384,128,0.02661866611904568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,64,0.015497777197096082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,16384,32,0.016154666741689045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,16384,0.15715732839372423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,65536,0.5753759808010525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,65536,0.5641857782999674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,16384,0.14899467097388372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,12288,0.12473155392540826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,12288,0.11515200138092041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,10240,0.09333511193593343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,10240,0.1040453314781189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,8192,0.0733777748213874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,8192,0.08406133121914333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,16384,0.1525528828303019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,7168,0.06543555524614122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,12288,0.12106222576565212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,7168,0.0758168896039327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,10240,0.11700622240702312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,8192,0.08825333250893487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,6144,0.057711111174689404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,6144,0.06581155459086101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,7168,0.08211733235253228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,5120,0.04878044459554884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,5120,0.05703644620047676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,6144,0.07135022348827787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,4096,0.039226667748557195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,4096,0.04725066820780436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,5120,0.06404533651140001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,3584,0.03500444359249539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,4096,0.05558133125305176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,3584,0.04361066553327772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,3072,0.03135022189882066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,65536,0.5449004703097874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,3072,0.03888622257444594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,3584,0.05517511235343086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,2560,0.026926222774717543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,3072,0.05080266793568929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,2560,0.03534488876660665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,2048,0.022521777285469904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,2560,0.04647466540336609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,2048,0.030851556195153132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,1536,0.017823111679818895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,2048,0.04324444466167026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,1536,0.026779555612140234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,1024,0.012240000069141388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,512,0.0081395556529363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,1536,0.0402800010310279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,1024,0.021874666213989258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,768,0.010258666343159145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,1024,0.03612800108061896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,768,0.02077155477470822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,128,0.004226666771703296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,768,0.0351528889603085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,512,0.01926488843229082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,512,0.031063109636306763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,256,0.005568888866239124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,256,0.016371554798550077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,256,0.027765333652496338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,128,0.015283554792404175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,64,0.003908444609906939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,12288,128,0.0258942229880227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,64,0.015541333291265698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,12288,32,0.00423911131090588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,12288,32,0.015767110718621146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,65536,0.49299022886488175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,16384,0.14290844069586858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,65536,0.5454177856445312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,16384,0.1420417759153578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,12288,0.09504799710379706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,10240,0.09735911422305638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,12288,0.1098711093266805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,10240,0.07899644639756945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,16384,0.14332711696624756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,8192,0.06396000252829658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,12288,0.11641866630978054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,8192,0.07818666431638929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,10240,0.10410488976372613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,7168,0.0572355588277181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,7168,0.07065155771043566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,8192,0.08465333117379083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,6144,0.050385776493284434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,6144,0.061520000298817955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,5120,0.042734222279654614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,5120,0.06117600202560425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,7168,0.07716000080108643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,5120,0.05277688966857063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,6144,0.06921333074569702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,4096,0.03390844331847297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,4096,0.045075555642445884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,65536,0.5224729114108616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,3584,0.03067822257677714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,4096,0.05366311139530606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,3584,0.04189777705404493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,3072,0.027388445205158655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,3072,0.0376142230298784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,3584,0.051256000995635986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,2560,0.023375999596383836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,3072,0.04902844296561348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,2560,0.03349955545531379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,2048,0.019701333509551156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,1024,0.010971555279360877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,2560,0.04420444369316101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,2048,0.029679109652837116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,1536,0.015106666419241162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,2048,0.04105777872933282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,1536,0.025431111454963684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,1536,0.03901155458556281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,1024,0.02138311167558034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,768,0.009448889229032729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,1024,0.0351528889603085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,768,0.02049422264099121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,512,0.00775022225247489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,768,0.03457244568400913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,512,0.017972444494565327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,256,0.005718222094906702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,512,0.032531556155946516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,256,0.01610844499535031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,128,0.004512000001139111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,256,0.02737066646416982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,32,0.01551999979548984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,128,0.015113777584499784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,64,0.003922666526503033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,10240,128,0.025624000363879736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,16384,0.09301244550281101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,65536,0.4136879973941379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,16384,0.10564977592892116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,10240,64,0.015214223000738354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,10240,32,0.004110222061475118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,65536,0.38959111107720273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,12288,0.11188711060418023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,12288,0.08277155293358697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,10240,0.06007377968894111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,10240,0.07297244336869982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,16384,0.11593777603573269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,8192,0.04930399854977926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,12288,0.0959048867225647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,8192,0.06005599763658312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,7168,0.05475111140145195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,7168,0.04395022326045566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,10240,0.11495021979014079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,8192,0.07455199956893921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,6144,0.039447112215889826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,6144,0.04768888817893135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,7168,0.06943022542529635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,65536,0.40107377370198566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,5120,0.03423733181423611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,5120,0.04146577914555868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,6144,0.06199466519885593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,4096,0.027621333797772724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,5120,0.05582311418321398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,4096,0.03730488816897074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,3072,0.030574222405751545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,3584,0.024783111280865137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,2560,0.0277128881878323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,3584,0.03395288851526048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,4096,0.049251556396484375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,3072,0.02188355558448368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,2048,0.024545777175161574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,3584,0.0469359987311893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,2560,0.0188426673412323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,3072,0.04387911160786947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,2048,0.015638222297032673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,2560,0.0403164459599389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,1536,0.011951111257076263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,1536,0.022188444932301838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,2048,0.03799644443723891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,1536,0.03651644455062018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,1024,0.009199110998047723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,1024,0.019440889358520508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,768,0.008045333127180735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,1024,0.033548444509506226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,768,0.01791288952032725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,512,0.0064399999876817065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,768,0.03323910964859857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,512,0.016192000773217943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,256,0.004297777596447202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,512,0.03067555692460802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,256,0.015100444356600443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,128,0.0037439999481042228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,256,0.026662222213215295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,128,0.013919111755159167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,8192,128,0.02493155499299367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,64,0.00350577798154619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,64,0.014149333039919535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,8192,32,0.0038337777886125776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,8192,32,0.01443733274936676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,65536,0.3397066593170166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,16384,0.08182399802737765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,65536,0.3594062328338623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,16384,0.10161866744359334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,12288,0.06381689177619086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,12288,0.08002666632334392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,10240,0.05390310949749417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,16384,0.13095288806491429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,10240,0.068904889954461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,12288,0.10492799679438274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,8192,0.04390044344796074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,7168,0.05297333333227369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,8192,0.058393776416778564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,10240,0.09313333034515381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,7168,0.039431111680136785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,65536,0.4755724271138509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,7168,0.07276978095372517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,8192,0.07820533381568061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,6144,0.03516799873775906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,6144,0.045963555574417114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,5120,0.029653333955340918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,5120,0.04120533333884345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,6144,0.07444266478220622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,4096,0.024540444215138752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,4096,0.03532444437344869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,3072,0.020250666472646926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,3584,0.04738400048679776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,5120,0.057714667585161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,3584,0.022287999590237934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,3584,0.03308444552951389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,4096,0.0503039989206526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,3072,0.029472000069088403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,2560,0.01723288827472263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,3072,0.044779555665122144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,2560,0.02720622221628825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,2048,0.013713777893119387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,2560,0.04070577687687344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,2048,0.024096000525686476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,2048,0.038076443804634943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,1536,0.011111111276679568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,1536,0.02206755512290531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,1024,0.008653333617581261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,1536,0.03635644581582811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,1024,0.01899111105336083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,1024,0.0338986681567298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,768,0.00741511086622874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,768,0.017143110434214275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,512,0.00628266649113761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,768,0.03330311179161072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,128,0.003726222034957674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,512,0.016013332539134555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,256,0.004274666723277834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,512,0.029708445072174072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,256,0.014904888139830695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,256,0.027042667071024578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,128,0.013966222604115805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,64,0.003641777982314428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,7168,128,0.025424000289705064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,64,0.013962666193644205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,7168,32,0.0037377778854635027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,7168,32,0.014137778017255994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,65536,0.29061510827806264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,16384,0.07292711072497897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,65536,0.35519199901156956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,12288,0.057822220855289035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,16384,0.09570666816499497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,12288,0.07479021946589152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,10240,0.04775822162628174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,16384,0.10163111156887478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,10240,0.0672195553779602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,12288,0.08235377735561795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,8192,0.0392595562669966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,8192,0.05477777785725064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,10240,0.0739884442753262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,7168,0.03544622328546312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,7168,0.05056800113783943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,8192,0.06492000155978732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,65536,0.3406337896982829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,6144,0.03147644466824002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,7168,0.06119910875956217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,5120,0.04976266622543335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,4096,0.03363555669784546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,6144,0.0440684457619985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,5120,0.02681066592534383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,6144,0.054551111327277295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,5120,0.03901244534386529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,4096,0.022091555926534865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,3584,0.02009600069787767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,4096,0.04399289025200737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,3584,0.031452443864610456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,3072,0.017822222577200997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,2048,0.012494222157531314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,3584,0.043007999658584595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,3072,0.02804533309406704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,1536,0.010249777800507015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,2048,0.03696533375316196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,3072,0.0407262212700314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,2560,0.01519733336236742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,2560,0.02571822206179301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,2560,0.03811466693878174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,768,0.007117333511511485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,2048,0.02318488889270359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,1536,0.021300445000330608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,1024,0.008031999899281396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,1536,0.035791112316979304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,1024,0.01906755566596985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,1024,0.0336791111363305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,768,0.016680000556839835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,768,0.03229066729545593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,512,0.005533333453867171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,512,0.01590577761332194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,256,0.0041617775956789655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,64,0.0034008889148632684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,128,0.025117332736651104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,512,0.028845333390765723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,256,0.014829332629839579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,128,0.003718222181002299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,6144,256,0.026384888423813715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,128,0.014010666145218743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,64,0.013818666338920593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,6144,32,0.003607999947335985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,6144,32,0.013944889108339945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,65536,0.23892090055677626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,65536,0.33237067858378094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,16384,0.07139111227459378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,16384,0.08811022175682916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,12288,0.05115111006630791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,12288,0.06771822108162774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,16384,0.12036532825893825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,10240,0.046021332343419395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,10240,0.05885599719153511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,12288,0.09680710898505317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,8192,0.03481866584883796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,8192,0.05077155431111654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,6144,0.028186665640936956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,10240,0.08653333452012803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,7168,0.03186310993300544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,65536,0.4360800054338243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,8192,0.07290044758054945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,7168,0.04672444529003567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,7168,0.06852977805667453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,6144,0.041152887874179415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,5120,0.024154666397306655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,6144,0.05961066484451294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,3584,0.018015111486117046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,5120,0.03640177845954895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,4096,0.019847111569510568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,5120,0.054487999942567616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,4096,0.031180444690916274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,4096,0.04681333237224155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,2560,0.013561777770519257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,3584,0.02935111191537645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,3072,0.01568000018596649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,2048,0.011369778050316704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,3584,0.045244445403416954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,3072,0.026399999856948853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,1536,0.010562666588359408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,3072,0.042371554507149585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,1024,0.007822222179836696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,2560,0.024414221445719402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,2560,0.03885511226124234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,2048,0.022142221530278523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,2048,0.03705244594150119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,1536,0.020492444435755413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,1536,0.035558223724365234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,1024,0.018380444910791185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,1024,0.03187111020088196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,768,0.006734222173690796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,768,0.016756445169448853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,512,0.005371555685997009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,768,0.03055822186999851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,512,0.015925332903862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,256,0.004092444562249714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,512,0.028432889117134943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,256,0.014656888114081489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,128,0.003624888757864634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,256,0.026349332597520616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,128,0.01386755539311303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,64,0.0033448889023727844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,5120,128,0.025011556016074285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,64,0.013766222529941134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,5120,32,0.0033902221669753394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,5120,32,0.013786666923099093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,65536,0.19428000185224745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,65536,0.3061093224419488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,16384,0.05657866928312513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,16384,0.08074044519000582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,12288,0.04597244328922696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,12288,0.0804613365067376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,16384,0.09720622168646918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,12288,0.06354488929112752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,10240,0.04029688901371426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,10240,0.055161777469846934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,65536,0.3093449009789361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,8192,0.033901333808898926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,8192,0.047503110435273915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,10240,0.08131555716196696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,7168,0.033322665426466204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,8192,0.06268711222542657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,7168,0.0431253347131941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,6144,0.02970400121476915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,6144,0.03684711125161912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,7168,0.059524443414476186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,5120,0.02018044392267863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,6144,0.05199377735455831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,5120,0.033482667472627424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,4096,0.015318221516079374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,5120,0.046465777688556247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,4096,0.028948442803488836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,3584,0.014031999640994601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,4096,0.040799111127853394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,3584,0.0270453327231937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,3072,0.012523555921183692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,3584,0.039911998642815485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,3072,0.02491377790768941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,2560,0.010728889041476779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,2560,0.023765333824687537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,3072,0.03745422098371718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,2048,0.00926755534278022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,2560,0.03535288903448317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,2048,0.02150844368669722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,1536,0.007704888780911763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,2048,0.0347360008292728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,1536,0.019924443629052904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,1024,0.006211555666393704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,768,0.030510223574108545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,512,0.00425511101881663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,1536,0.03383288780848185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,1024,0.017473777135213215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,768,0.0052639999323421055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,1024,0.030894223186704848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,256,0.0262435558769438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,768,0.0165928883685006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,512,0.015608888533380298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,256,0.003677333394686381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,512,0.027426666683620874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,32,0.003411555455790626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,256,0.014753777119848462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,128,0.0033644443998734155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,128,0.01366311146153344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,4096,128,0.024519999821980793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,4096,64,0.003141333245568805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,64,0.013668444421556262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,4096,32,0.013626666532622443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,65536,0.17192444536421034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,65536,0.28880710071987575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,16384,0.04983289043108622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,16384,0.09530133008956909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,16384,0.07807022333145142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,12288,0.04159644577238295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,12288,0.0786844425731235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,12288,0.06075644493103027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,10240,0.03628177775277032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,10240,0.053264889452192515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,65536,0.30367999606662327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,8192,0.024736000431908503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,8192,0.0624613364537557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,7168,0.040828443235821195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,10240,0.07073244121339586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,8192,0.045751998821894325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,7168,0.022619555393854778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,6144,0.020266667008399963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,7168,0.057406220171186656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,6144,0.03584533267550998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,5120,0.017094223035706412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,6144,0.05024622215165032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,5120,0.03244444396760728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,4096,0.013896889156765409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,5120,0.04552177919281853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,4096,0.02838488750987583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,3584,0.012663111090660095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,4096,0.04062666826777988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,3584,0.026777777406904433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,3072,0.011124444504578909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,3584,0.03922755519549052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,3072,0.024711999628278945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,2560,0.009918221996890174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,1536,0.007408888803588019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,3072,0.03762933280732896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,2560,0.02316355539692773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,2560,0.035143110487196184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,2048,0.00849955528974533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,2048,0.021298666795094807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,2048,0.034272889296213783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,1536,0.01975733372900221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,1024,0.005966222120655908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,1536,0.0335457788573371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,1024,0.01716711123784383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,768,0.004453333301676644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,1024,0.03064889046880934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,768,0.016351110405392118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,512,0.003962666624122196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,768,0.02940088841650221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,512,0.01576266686121623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,256,0.003488000068399641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,128,0.02423644396993849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,512,0.027387556102540758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,256,0.014583110809326172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,32,0.013726222018400827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,128,0.0032248888164758682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3584,256,0.025757332642873127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,128,0.013721778160995908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,64,0.003297777846455574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3584,64,0.01367022262679206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3584,32,0.003318222239613533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,65536,0.15218844678666857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,65536,0.27918844752841526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,65536,0.3994764486948649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,16384,0.044376889864603676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,16384,0.07585510942671034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,12288,0.037195556693606906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,16384,0.11368000507354736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,12288,0.057961775196923145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,10240,0.032552000549104475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,7168,0.024022221565246582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,12288,0.08928355243470933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,10240,0.051510221428341336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,8192,0.030555556217829388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,7168,0.06340799729029338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,10240,0.07842399676640828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,8192,0.0437520013915168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,8192,0.0672631131278144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,7168,0.039566222164365984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,4096,0.017110221915774874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,6144,0.02494222256872389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,6144,0.0351582235760159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,5120,0.01908266709910499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,6144,0.055211553970972695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,5120,0.03158844510714213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,5120,0.04969066712591383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,3584,0.04248799880345663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,4096,0.0277146663930681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,3584,0.01165244479974111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,4096,0.044228444496790566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,3584,0.02608266638384925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,3072,0.010416888528399997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,3072,0.024233778317769367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,2560,0.009287111461162567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,3072,0.04001599881384108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,2560,0.023018666439586218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,2560,0.0369742214679718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,2048,0.007988444632954067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,2048,0.021187555458810594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,1536,0.007063999772071838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,2048,0.03597066799799601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,1536,0.019175999694400363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,1024,0.005561777700980504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,1536,0.033415112230512835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,1024,0.017122666041056316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,768,0.004488000025351842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,1024,0.02923911147647434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,768,0.01643022232585483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,512,0.003916444463862313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,768,0.029213332467608984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,512,0.01570044457912445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,256,0.003469333259595765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,512,0.028121779362360638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,256,0.014493332968817817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,128,0.0031893334041039148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,256,0.025955556167496577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,128,0.013772444592581855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,64,0.003148444410827425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,3072,128,0.02443733314673106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,64,0.013619555367363824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,3072,32,0.0032551110618644287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,16384,0.07120977507697211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,3072,32,0.013516444298956128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,12288,0.0317831105656094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,65536,0.13087022304534912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,12288,0.056458665264977344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,65536,0.2661457856496175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,10240,0.049936887290742665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,16384,0.039119998613993325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,16384,0.08038666513231066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,10240,0.028598222467634413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,12288,0.06774399677912395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,8192,0.05576799975501167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,65536,0.2937297821044922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,8192,0.02439466615517934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,10240,0.06163377894295586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,8192,0.042487111356523305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,7168,0.024659555819299486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,7168,0.038943999343448214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,6144,0.020540444387329947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,4096,0.01296622223324246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,7168,0.052729778819613986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,6144,0.03475288881195916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,5120,0.01867999964290195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,6144,0.04642577634917366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,5120,0.031457778480317854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,3072,0.010919111470381418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,5120,0.04199377695719401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,4096,0.02774222195148468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,4096,0.038093331787321306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,3584,0.012234666281276278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,3584,0.0262773334980011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,3584,0.037343111303117536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,3072,0.02421422302722931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,3072,0.03564533260133531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,2560,0.00871288859181934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,2560,0.02265600032276577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,2048,0.007740444607204861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,2560,0.033719112475713096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,2048,0.01997422178586324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,1536,0.006704888823959563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,768,0.004399111287461387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,2048,0.030807110998365615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,1536,0.018301332990328472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,1536,0.031319111585617065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,1024,0.005028444445795483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,1024,0.016846223009957206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,1024,0.028196444114049275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,768,0.016100444727473788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,512,0.003916444463862313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,768,0.02829066581196255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,512,0.015320888823933072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,256,0.003458666718668408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,128,0.02403733299838172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,512,0.02702933384312524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,256,0.01425866617096795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,128,0.0031439999325407874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2560,256,0.02573777735233307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,128,0.013495999905798169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,64,0.002983111060327954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,64,0.013612444202105204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2560,32,0.0030533334033356774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2560,32,0.01352355546421475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,65536,0.10263999965455796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,65536,0.26492267184787327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,16384,0.033237334754731916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,16384,0.07104888889524671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,12288,0.027109333210521277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,10240,0.04961511161592272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,16384,0.07949155569076538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,12288,0.056136886278788246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,65536,0.30056799782647026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,10240,0.02425866656833225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,12288,0.06739377975463867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,8192,0.023557333482636347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,10240,0.06072533130645752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,8192,0.04212799999448988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,8192,0.053617778751585216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,7168,0.02106399999724494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,7168,0.03879377908176846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,7168,0.050851556989881724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,6144,0.01828888886504703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,6144,0.03478488988346524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,5120,0.01593155496650272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,6144,0.04487644301520454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,3584,0.012224888636006249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,5120,0.031000001562966242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,5120,0.041229334142473005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,4096,0.013190222283204397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,4096,0.027485332555241052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,4096,0.03773422373665704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,3584,0.02574133376280467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,3072,0.010915555887752108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,3584,0.03715377714898851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,3072,0.02408355474472046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,2560,0.00833422193924586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,2560,0.022491556074884202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,3072,0.03565333286921183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,2048,0.007304000357786815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,1024,0.004503111044565837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,1536,0.029881778690550063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,2560,0.03288622366057502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,768,0.00434044458799892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,2048,0.01979200045267741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,768,0.015974221958054435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,1536,0.006380444599522485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,2048,0.032095111078686185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,1536,0.0181013329161538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,1024,0.016713778177897137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,1024,0.028120001157124836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,512,0.0037573331760035623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,768,0.028175112273957994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,512,0.015121777852376303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,512,0.02716355522473653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,64,0.0029306666304667792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,256,0.00332266671790017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,256,0.014189332723617554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,128,0.003030222323205736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,256,0.025498666697078284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,128,0.013448889056841532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,2048,128,0.02405866649415758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,65536,0.26181422339545357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,16384,0.02752711044417487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,16384,0.07087644603517321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,64,0.01346666696998808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,2048,32,0.003165333428316646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,2048,32,0.013335111240545908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,65536,0.08110311296251085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,12288,0.022525333695941504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,65536,0.2547084490458171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,16384,0.07430489195717706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,12288,0.05623910824457804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,10240,0.0258942229880227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,12288,0.06195466385947334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,10240,0.04933155576388041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,8192,0.02123733361562093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,10240,0.056222220261891685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,8192,0.04142133394877116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,7168,0.018805333309703402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,8192,0.049794667296939425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,7168,0.03840355409516229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,6144,0.016917333006858826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,4096,0.012327999704413943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,7168,0.04761955473158094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,6144,0.034268445438808866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,5120,0.014779556128713818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,6144,0.04284888837072584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,5120,0.03092977735731337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,5120,0.040264887942208186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,4096,0.027436445156733196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,4096,0.03626044591267904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,2560,0.007663999994595845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,3584,0.01128888875246048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,2560,0.02113955550723606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,3584,0.025939555631743536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,3072,0.010122666756312052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,3584,0.035763555102878146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,3072,0.023894222246276006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,3072,0.034677333301968045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,2048,0.006679111056857639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,2560,0.03289066751797994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,2048,0.019482667247454327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,1536,0.005489777773618698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,2048,0.030400888787375555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,1536,0.018024888303544786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,1024,0.004507555729813046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,1536,0.02979466650221083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,1024,0.01657777859105004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,1024,0.02773422168360816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,768,0.004055110944641961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,768,0.015992000699043274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,768,0.0280693338976966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,512,0.0036515556275844574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,512,0.015174221661355762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,256,0.0032239999208185407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,512,0.02714666724205017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,256,0.014074666632546319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,128,0.003034666594531801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,256,0.02517066730393304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,128,0.013375110924243927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1536,128,0.023945776952637568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,64,0.0028417776856157514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,64,0.013419555293189155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,16384,0.02144977781507704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1536,32,0.0029760001020299066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1536,32,0.013431111143694984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,65536,0.058133330610063344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,65536,0.26318756739298504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,16384,0.07134044170379639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,12288,0.020790222618314955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,65536,0.3393893241882324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,16384,0.09548711114459568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,12288,0.05570489168167114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,10240,0.018043556147151522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,12288,0.07888088623682658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,10240,0.04863111178080241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,8192,0.015542222393883599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,10240,0.07125244537989299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,8192,0.041351109743118286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,7168,0.013737777868906656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,8192,0.0602088901731703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,7168,0.03828444414668613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,6144,0.012291555603345236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,6144,0.033647110064824425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,7168,0.057108441988627114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,5120,0.011162666810883416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,6144,0.049469331900278725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,5120,0.03051111102104187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,4096,0.009477333062224919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,3072,0.008114666574531132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,5120,0.04583200150065952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,4096,0.027062222361564636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,3584,0.00870311094654931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,4096,0.04049866729312473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,3584,0.024324445260895625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,3584,0.03863111138343811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,3072,0.022481777601771887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,2560,0.007244444555706448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,3072,0.03679555654525757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,2560,0.021168000168270532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,2048,0.0061928890645504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,2560,0.03317155440648397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,1536,0.031508445739746094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,2048,0.019367999500698514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,1536,0.005372444374693765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,2048,0.032419555717044406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,1536,0.017994667092959087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,1024,0.004497777670621872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,1024,0.01647911138004727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,768,0.004056888735956616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,256,0.0031928889867332247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,768,0.015823110938072205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,1024,0.028905779123306274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,128,0.0029297777348094513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,512,0.003612444632583194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,768,0.02882844540807936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,512,0.015058666467666626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,512,0.027223111854659185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,256,0.014174222946166992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,256,0.025635555386543274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,128,0.013250666360060373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,64,0.002846222163902389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,1024,128,0.024177778098318312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,64,0.013378666506873237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,1024,32,0.0028417776856157514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,1024,32,0.013240888714790344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,65536,0.04725511206520928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,65536,0.2643306785159641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,16384,0.022965333527988855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,10240,0.015955555770132277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,16384,0.07061777512232463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,65536,0.30169688330756295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,12288,0.01792711185084449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,16384,0.091776000128852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,8192,0.04136711027887132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,12288,0.055508444706598915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,12288,0.07517866955863105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,10240,0.0479964448346032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,8192,0.01296622223324246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,10240,0.06648533211814033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,7168,0.012970666918489667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,8192,0.0584204461839464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,7168,0.038374221987194486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,6144,0.0120000003112687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,7168,0.05563555823432075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,6144,0.03359199894799127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,5120,0.010593777729405297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,6144,0.04828088813357883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,5120,0.029446221060223047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,4096,0.009032888544930352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,5120,0.04500000013245476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,4096,0.025948445002237957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,3584,0.008455110920800103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,4096,0.0389048887623681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,3584,0.024259555670950148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,3072,0.007594666547245449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,3584,0.03692888882425096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,3072,0.0226275556617313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,2560,0.006766222417354584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,3072,0.03503733211093479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,2560,0.021006221572558086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,2048,0.005602666487296422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,2560,0.03353777858946059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,2048,0.019489778412712943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,1536,0.004918222212129169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,2048,0.03176177872551812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,1536,0.018132444885041978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,768,0.016007110476493835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,1024,0.0041928887367248535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,1536,0.030935111973020766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,1024,0.016580444243219163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,768,0.0038577777643998465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,1024,0.028423110644022625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,768,0.028343998723559912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,512,0.0035013332962989807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,512,0.015052444405025907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,256,0.0031644445326593188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,512,0.02702577743265364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,256,0.014239999983045789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,128,0.0029120000286234748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,256,0.02558044426971012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,128,0.013468444347381592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,768,128,0.024076445235146418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,64,0.002732444347606765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,16384,0.01536711057027181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,64,0.013434666726324292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,768,32,0.00282222218811512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,12288,0.01477777792347802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,768,32,0.013394667042626275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,65536,0.03784444597032335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,65536,0.26495467291937935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,16384,0.06993689139684041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,65536,0.22580356068081328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,16384,0.0692968898349338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,12288,0.054550220568974815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,10240,0.01330044451687071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,12288,0.05693511168162028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,10240,0.04771555463473002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,10240,0.05182488759358724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,6144,0.010257778068383535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,7168,0.0449022220240699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,8192,0.011599111060301462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,5120,0.009308444129096137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,8192,0.04093599981731839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,7168,0.010904889139864178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,8192,0.04683377676539951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,7168,0.037680887513690524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,6144,0.032495998673968844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,6144,0.04097155398792691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,5120,0.02908088763554891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,5120,0.037652442852656044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,4096,0.008536000218656328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,4096,0.025320000118679468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,4096,0.033082667324278094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,3584,0.007883555359310573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,3584,0.024117334021462336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,3072,0.007089777953094906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,2560,0.03007022208637661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,3584,0.03243822190496657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,3072,0.022364444202846948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,3072,0.03155466583040025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,2560,0.006166222194830577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,2560,0.020761777957280476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,2048,0.005414222263627582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,2048,0.019039111004935372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,2048,0.02970044480429755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,1536,0.0047973332305749255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,1536,0.017493334081437852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,1536,0.028992888000276353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,1024,0.0041031113101376426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,1024,0.01644000079896715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,768,0.0037600000699361167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,1024,0.027311111489931744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,768,0.015942222542232938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,768,0.027759111589855615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,512,0.00342399999499321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,512,0.014907555447684394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,256,0.0030542222989930045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,512,0.026672000686327618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,256,0.013864888913101621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,128,0.0028888889484935333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,256,0.025244444608688354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,128,0.01330044451687071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,64,0.0026942222482628296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,512,128,0.023751111494170293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,64,0.013351110948456658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,512,32,0.0027715555495686005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,512,32,0.01312355531586541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,65536,0.03231288989384969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,65536,0.2649537722269694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,16384,0.01144266708029641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,65536,0.225038210550944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,16384,0.06893599695629544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,12288,0.010067555639478896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,16384,0.06844355662663777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,12288,0.05415022373199463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,12288,0.05654311180114746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,10240,0.009187555147541894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,10240,0.0470106667942471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,10240,0.051348444488313466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,8192,0.008393777741326226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,8192,0.0396453340848287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,8192,0.04622844523853726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,7168,0.0074924445814556545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,7168,0.03623111049334208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,7168,0.04479111234347025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,6144,0.007176000210973952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,6144,0.03194311261177063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,6144,0.039312889178593956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,5120,0.00721599989467197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,5120,0.02845777736769782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,4096,0.007091555330488417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,5120,0.03557599915398492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,4096,0.026591999663247004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,4096,0.03250844279925028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,3584,0.0070062221752272705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,3584,0.023465777436892193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,3072,0.006706666615274217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,3584,0.03187733226352268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,3072,0.02199111051029629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,2560,0.006036444256703059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,3072,0.030976000759336684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,2560,0.020247111717859905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,2560,0.029269334342744615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,2048,0.005371555685997009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,2048,0.019023999571800232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,2048,0.028970665401882593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,1536,0.0047333331571684945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,1536,0.01737600068251292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,1536,0.028789334826999243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,1024,0.004032000071472592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,1024,0.015980444020695157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,768,0.0037128888070583344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,1024,0.027012444204754297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,768,0.015281778242852954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,512,0.003400000019205941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,768,0.02737866673204634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,512,0.014578666951921252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,512,0.026579555537965562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,256,0.002999111182159848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,256,0.013912889692518445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,128,0.0028168888141711554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,256,0.02515911062558492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,128,0.013230221966902414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,256,128,0.023702222439977858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,64,0.0026853332916895547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,64,0.013238222234778933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,256,32,0.0026968889352348116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,256,32,0.013236444857385425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,65536,0.023158222436904907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,65536,0.26529423395792645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,16384,0.00889866633547677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,65536,0.22463289896647134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,16384,0.06893155309889051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,12288,0.007804444266690149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,16384,0.06763555606206258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,12288,0.052998221582836576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,12288,0.05606489049063789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,10240,0.007134222322040134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,10240,0.046337776713901095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,10240,0.05025510986646017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,8192,0.007002666592597961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,6144,0.031735999716652766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,8192,0.038745777474509344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,7168,0.006807999892367258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,8192,0.04467822114626566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,7168,0.035445332527160645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,6144,0.0064693333374129395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,7168,0.042479998535580106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,5120,0.0069066666894488865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,6144,0.03831911087036133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,5120,0.028035554620954726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,4096,0.006480888773997624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,5120,0.035820444424947105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,4096,0.024911110599835713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,3072,0.031110223796632554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,3584,0.006892444358931647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,2560,0.02033866610791948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,4096,0.032167110178205706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,3584,0.023510222633679707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,3584,0.03187911046875848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,3072,0.006583999842405319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,3072,0.021663111117151048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,2560,0.006001777946949005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,2048,0.005323555734422471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,2560,0.029037333197063867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,2048,0.018730666902330186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,2048,0.02898933490117391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,1536,0.004607111215591431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,1536,0.017062221964200337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,1536,0.028299556838141546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,1024,0.00397777764333619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,1024,0.015849777393870883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,768,0.0036071112586392295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,1024,0.026936888694763184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,768,0.015248000621795654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,512,0.0033324443631701996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,768,0.027433777848879497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,256,0.024825778272416856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,512,0.014376888672510782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,256,0.0029786665820413162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,512,0.026103110777007207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,256,0.01370666672786077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,128,0.0027342221389214196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,32,0.013017777767446307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,128,0.013111111190583972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,160,128,128,0.023598222268952265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,64,0.0026337777574857077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,128,64,0.01313333296113544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,128,32,0.002616000051299731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,65536,0.013759111364682516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,10240,0.00685511115524504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,16384,0.007053333024183909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,65536,0.26508355140686035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,16384,0.0684666633605957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,12288,0.006837333242098491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,12288,0.05323822299639384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,8192,0.006671111202902264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,10240,0.04507199923197428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,7168,0.006588444527652528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,8192,0.03846222162246704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,7168,0.035375111632876925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,6144,0.0063546668324205614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,6144,0.031473779016070895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,5120,0.006783111227883234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,5120,0.028104000621371802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,4096,0.006319111006127463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,4096,0.024970667229758367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,3584,0.006763555523422029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,3584,0.0232933329211341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,3072,0.006503110958470239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,3072,0.02164799968401591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,2560,0.005961777849329843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,2560,0.020223110914230347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,2048,0.005249777601824866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,1536,0.004590222405062782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,2048,0.01868088874551985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,1536,0.01715377800994449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,1024,0.003929777691761653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,1024,0.01565511193540361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,768,0.003607999947335985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,768,0.015079110860824585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,512,0.003240888938307762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,512,0.014367111855083041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,256,0.0032382222513357797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,256,0.013659555878904132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,128,0.0027422221998373666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,128,0.01313244468635983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,64,0.002612444468670421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,64,0.013011555704805585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,64,32,0.002605333303411802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,64,32,0.012863111164834766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,65536,0.013280889226330651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,16384,0.006792888873153263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,65536,0.2652897834777832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,16384,0.06777599785063002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,12288,0.006923555499977536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,12288,0.05183555682500204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,10240,0.006700444552633498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,10240,0.045406222343444824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,8192,0.006675555474228329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,8192,0.03863822089301215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,7168,0.006385777973466449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,7168,0.0351893338892195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,6144,0.006305777778228124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,5120,0.006641777853171031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,6144,0.03135022189882066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,5120,0.028002666102515325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,4096,0.006253333141406377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,4096,0.024953777591387432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,3584,0.006684444430801604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,3584,0.023206222388479445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,3072,0.0064088888466358185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,3072,0.021632888250880774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,2560,0.005960888746711943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,2560,0.02014755540423923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,2048,0.0052560000783867305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,2048,0.01868355605337355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,1536,0.004589333302444882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,1536,0.017064889272054035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,1024,0.003972444269392225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,1024,0.015767999821239047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,768,0.003569777641031477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,768,0.015096000499195524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,512,0.0032435556252797446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,512,0.014471999473041959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,256,0.0029306666304667792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,256,0.013537777794731988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,128,0.0027022223091787765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,128,0.013136000268989138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,64,0.002587555597225825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,64,0.013011555704805585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,160,32,32,0.00258577780591117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,160,32,32,0.012934221989578672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,16384,0.39487287733289933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,16384,0.7303297784593371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,12288,0.3077475494808621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,12288,0.5574400160047743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,10240,0.46409331427680117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,10240,0.2635262277391222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,8192,0.20880533589257133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,8192,0.3753004339006212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,7168,0.32646221584743923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,7168,0.20815466509924996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,6144,0.2852444383833143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,6144,0.16214133633507624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,5120,0.2516169018215603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,5120,0.13728445106082493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,4096,0.19341244962480333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,4096,0.1134951114654541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,8192,0.2525386545393202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,12288,0.43203022744920516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,16384,0.43187999725341797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,3584,0.1693253384696113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,3584,0.1052302254570855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,3072,0.14717333846622044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,3072,0.09042755762736003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,10240,0.30938487582736546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,2560,0.12447910838656956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,4096,0.14450311660766602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,2560,0.07778666416803996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,2048,0.10143111149470012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,6144,0.19779733816782633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,7168,0.2247920036315918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,2048,0.0745368864801195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,1536,0.0790319972568088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,5120,0.17060711648729113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,1536,0.05506755577193367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,1024,0.05583733320236206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,1024,0.043826666143205434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,768,0.049470222658581205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,768,0.03792000148031447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,512,0.0325911111301846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,512,0.032130665249294706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,512,0.048152890470292836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,256,0.019376888871192932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,256,0.02502222193611993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,768,0.05437155564626058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,128,0.011676444775528379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,128,0.0206995556751887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,2560,0.10308621989356147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,3584,0.12440355618794759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,64,0.013439999686347114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,64,0.02386844489309523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,65536,32,0.01387733303838306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,65536,32,0.02439111140039232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,65536,0.4491306410895453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,65536,0.7581493589613172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,1536,0.07319377528296576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,16384,0.18346756034427217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,16384,0.12449599636925592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,12288,0.13896711667378744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,12288,0.09940799739625718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,3072,0.11877866586049397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,10240,0.11680444081624348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,2048,0.08461866776148479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,10240,0.08595822254816692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,1024,0.06245599852667915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,8192,0.09380710787243313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,8192,0.06982400019963582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,7168,0.08247377475102742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,7168,0.06295555830001831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,256,0.03917600048912896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,6144,0.07304622067345513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,6144,0.056736886501312256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,65536,128,0.03352266550064087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,5120,0.06135555770662096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,5120,0.04907200071546766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,12288,0.12818577554490831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,16384,0.1646933290693495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,4096,0.0502800014283922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,4096,0.04626577761438158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,3584,0.044864889648225575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,3584,0.0388062232070499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,8192,0.08684800068537395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,65536,0.5116311179267036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,10240,0.10353866550657485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,3072,0.03959733247756958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,3072,0.03573333223660787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,2560,0.0336097776889801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,2560,0.03244444396760728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,2048,0.028802666399214003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,2048,0.029823111163245306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,7168,0.07891199986139934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,1536,0.022426666484938726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,1536,0.024412444896168176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,6144,0.07268000311321683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,1024,0.01567555632856157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,1024,0.02052622205681271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,5120,0.064264886909061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,768,0.01238844460911221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,768,0.0195768889453676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,3584,0.054959111743503146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,4096,0.05765689081615872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,512,0.00927111092540953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,512,0.01812355551454756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,2560,0.04843644301096598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,256,0.00675022229552269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,256,0.014675555957688225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,3072,0.051049778858820595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,128,0.004605333424276776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,128,0.013800889253616333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,64,0.004249777644872665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,64,0.014367111855083041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,1536,0.040263109736972384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,16384,32,0.004652444273233414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,16384,32,0.014434667097197639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,2048,0.04418666826354133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,65536,0.5503093401590983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,65536,0.39075377252366805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,16384,0.1355946726269192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,16384,0.11663377285003662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,12288,0.10448355807198419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,12288,0.08606311347749497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,1024,0.034958223501841225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,10240,0.08786400159200032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,10240,0.07444800270928277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,768,0.03435822327931722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,8192,0.0707448919614156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,8192,0.06095288859473335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,512,0.033302221033308245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,7168,0.062365333239237465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,7168,0.05469333463244968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,256,0.02908710969818963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,6144,0.055027554432551064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,16384,128,0.02494844463136461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,6144,0.048919111490249634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,5120,0.046867556042141385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,5120,0.04282755653063456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,16384,0.1361653274959988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,4096,0.038146668010287814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,12288,0.10590577787823147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,4096,0.03712266683578491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,3584,0.03414933217896356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,3584,0.03408088948991563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,65536,0.46389245986938477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,3072,0.030280888080596924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,3072,0.03110755483309428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,10240,0.0912097758717007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,2560,0.02590755621592204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,2560,0.028241776757770117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,8192,0.07675288783179389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,2048,0.021672889590263367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,2048,0.026263111167483862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,6144,0.06418222188949585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,5120,0.05800888935724894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,7168,0.07090666559007433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,1536,0.01735911104414198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,1536,0.023715555667877197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,1024,0.01204622205760744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,1024,0.01977777812216017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,768,0.009740444521109263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,768,0.018785778019163344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,3584,0.04996800091531542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,512,0.007732444339328342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,512,0.01716711123784383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,4096,0.05242400036917793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,256,0.005299555758635203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,256,0.014545778433481852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,3072,0.04703199863433838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,128,0.0041511112617121804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,128,0.013963555296262106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,2560,0.044050667021009654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,64,0.0038000001675552795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,64,0.014106666048367819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,12288,32,0.004238222208287981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,12288,32,0.014260444376203748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,65536,0.47661511103312176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,65536,0.36077332496643066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,2048,0.03933066791958279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,768,0.03344000048107571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,16384,0.1147528886795044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,16384,0.10064088635974461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,12288,0.08773244751824273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,12288,0.0802897744708591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,1536,0.037504888243145414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,10240,0.07449599769380358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,10240,0.07008888986375597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,1024,0.03341866532961527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,8192,0.05991911225848728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,8192,0.057483553886413574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,256,0.026234666506449383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,128,0.024663999676704407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,7168,0.05309333403905233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,7168,0.05177866750293308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,6144,0.04738577869203356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,6144,0.05010133319430881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,12288,512,0.03176000052028232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,5120,0.039805332819620766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,5120,0.04085155659251743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,16384,0.11489333046807183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,4096,0.032744000355402626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,4096,0.03550666570663452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,65536,0.4188240104251438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,12288,0.09519644578297932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,3584,0.029145777225494385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,3584,0.0326186650329166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,3072,0.02597066594494714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,10240,0.08474844694137573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,3072,0.029503110382292006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,2560,0.022295999858114455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,2560,0.02652799917591943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,8192,0.07142666975657146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,2048,0.01843377782238854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,2048,0.024541333317756653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,7168,0.06619733572006226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,1536,0.014350222216712104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,1536,0.021088000800874498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,6144,0.06034488810433281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,1024,0.010378667049937779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,1024,0.01940711173746321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,5120,0.054775999652014844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,768,0.008794666992293464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,768,0.01809955636660258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,3584,0.04820177621311612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,512,0.0069431112044387394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,512,0.015471999843915304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,4096,0.04925688770082262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,256,0.004658666749795278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,256,0.014449778530332776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,2560,0.04216622312863668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,128,0.00396799999806616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,128,0.013551111022631327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,3072,0.04424622323777941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,64,0.003724444243643019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,768,0.03268622358640035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,64,0.014037332601017423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,10240,32,0.004097777936193678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,10240,32,0.01408266690042284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,65536,0.40177422099643284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,65536,0.27523554695977104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,16384,0.10339644220140244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,16384,0.08040710952546862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,1536,0.03596177697181702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,12288,0.07138399945365058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,12288,0.06546577480104235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,2048,0.037956443097856306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,10240,0.060115555922190346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,10240,0.05733955568737454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,1024,0.0329422222243415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,8192,0.04855111241340637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,8192,0.04695822132958306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,128,0.024121777878867254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,7168,0.042876445584827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,7168,0.04266311062706841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,256,0.025756445195939805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,6144,0.038586666186650596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,6144,0.03885066509246826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,16384,0.1024071110619439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,5120,0.032595554987589516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,65536,0.353075557284885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,5120,0.03443466623624166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,4096,0.02697066631582048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,4096,0.030456887351142034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,12288,0.08580178022384644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,10240,512,0.03024266825781928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,3584,0.0240657776594162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,3584,0.02812355425622728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,3072,0.021367111139827307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,3072,0.025213332639800176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,10240,0.07504266500473022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,2560,0.01829600003030565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,2560,0.02307555576165517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,8192,0.0638257794910007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,2048,0.01519466605451372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,2048,0.02115733259254032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,7168,0.05903999672995674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,1536,0.011769777370823754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,1536,0.019345778557989333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,6144,0.05369244350327385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,1024,0.008902221918106079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,1024,0.018198221921920776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,5120,0.04942399925655789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,768,0.007606222397751278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,768,0.016599110431141324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,4096,0.045174221197764076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,512,0.006234666539563074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,512,0.014367999302016364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,3584,0.04294311006863912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,256,0.0040737779604064096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,256,0.01367911116944419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,3072,0.04001511136690775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,128,0.003629333443111844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,128,0.012940444052219391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,2560,0.03834755553139581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,64,0.003350222276316749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,64,0.013256000147925483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,8192,32,0.0036186666952239144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,8192,32,0.013285333083735572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,65536,0.3494417667388916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,65536,0.25512801276312935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,2048,0.035707556539111666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,16384,0.09140621953540379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,16384,0.07548266649246216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,1536,0.03482666611671448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,12288,0.0724560022354126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,12288,0.06109155548943413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,1024,0.03147466646300422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,10240,0.05389777819315592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,10240,0.05351822243796455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,256,0.025220443805058796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,8192,0.042734222279654614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,512,0.027873777680926855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,128,0.023768888579474554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,8192,0.04366222355100843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,7168,0.03887466589609782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,7168,0.040051556295818753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,6144,0.03450311223665873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,8192,768,0.032293332947625056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,6144,0.03668444355328878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,5120,0.029264890485339697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,5120,0.03245688809288873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,65536,0.31873422198825413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,4096,0.023873777853118047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,4096,0.028219554159376357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,16384,0.09734133217069839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,3584,0.021335111724005804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,3584,0.026343110534879897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,12288,0.08006488614612155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,3072,0.018918222851223417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,3072,0.024145777026812237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,10240,0.07150933477613661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,2560,0.016335111525323655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,2560,0.022572444544898138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,8192,0.06137510803010729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,2048,0.01331733332739936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,7168,0.05707822243372599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,2048,0.020675554871559143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,1536,0.010566222170988718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,6144,0.05176800158288744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,1536,0.01889333294497596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,1024,0.008177777959240807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,5120,0.047500444783104785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,1024,0.017701332767804463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,768,0.006976000136799282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,768,0.015016888578732809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,4096,0.043262223402659096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,512,0.005741333381997214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,512,0.014379555980364481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,3584,0.041342222028308444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,256,0.0038319999972979226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,256,0.013787555197874704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,3072,0.038254221280415855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,128,0.0034053333931499054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,128,0.012999110751681857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,2560,0.03756533397568597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,64,0.0031902222997612427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,64,0.013215111361609565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,7168,32,0.0034808889031410217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,7168,32,0.013144888811641268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,65536,0.3105013370513916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,1536,0.03447377681732178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,768,0.03188088867399428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,65536,0.23668622970581055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,16384,0.07831022474500868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,2048,0.03490399983194139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,16384,0.06978844271765815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,1024,0.03200177682770623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,12288,0.05466933382882012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,12288,0.056741330358717174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,10240,0.0466648903157976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,10240,0.050182223320007324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,8192,0.0377200014061398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,8192,0.04116888840993246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,512,0.02787999974356757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,7168,0.033306668202082314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,7168,0.03793955511516995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,256,0.025543111893865798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,6144,0.02944533361328973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,6144,0.034486220942603216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,7168,128,0.02410488824049632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,5120,0.025270221961869135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,5120,0.03028977910677592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,16384,0.09763733545939128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,4096,0.021012443635198805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,12288,0.08137422137790255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,10240,0.07265066438251071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,4096,0.026353778110610113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,3584,0.018757333358128864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,3584,0.024586665961477492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,65536,0.30249065823025173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,3072,0.016549333930015564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,8192,0.061344888475206166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,3072,0.022527111901177302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,2560,0.014185777968830533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,2560,0.021224000387721594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,2048,0.011533333195580376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,2048,0.019837333096398246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,7168,0.05769599808586968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,1536,0.009450666606426239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,1536,0.0184888889392217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,6144,0.053017778529061206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,1024,0.007426666716734569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,1024,0.017091555727852713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,5120,0.04741777645217048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,768,0.006430222342411677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,768,0.014679999815093147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,4096,0.04204977883232964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,3584,0.04092177748680115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,512,0.0046737777690092725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,512,0.014339556296666464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,256,0.003629333443111844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,256,0.013565333353148567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,3072,0.03743911120626662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,128,0.0032906666811969546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,2560,0.03684799869855245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,128,0.012916444076432122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,64,0.003067555526892344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,2048,0.034325334760877825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,64,0.013021333350075616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,6144,32,0.0033191111352708605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,6144,32,0.012989333106411828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,65536,0.25366756651136607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,65536,0.22270311249627006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,16384,0.0670800010363261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,16384,0.06544799937142266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,1536,0.033831109603246055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,12288,0.054142223464118115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,12288,0.05351288782225715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,768,0.028493334849675495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,10240,0.04662400152948168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,10240,0.047255999512142606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,1024,0.03109333250257704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,8192,0.03165600034925673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,8192,0.038929777012930974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,256,0.025045333637131587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,128,0.02360711163944668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,7168,0.028891556792789038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,7168,0.03591377867592706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,6144,0.02958222230275472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,6144,0.03203466534614563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,6144,512,0.026945778065257605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,5120,0.02182666626241472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,5120,0.02840977907180786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,16384,0.08553866545359294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,4096,0.018068444397714403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,4096,0.025014221668243408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,65536,0.2590693367852105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,3584,0.01608000033431583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,3584,0.023188443647490606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,12288,0.0706826647122701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,3072,0.014168888330459595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,3072,0.02201333310869005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,10240,0.0639431079228719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,2560,0.011855999628702799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,2560,0.020502222908867728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,8192,0.054767111937204994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,2048,0.01017777787314521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,2048,0.01980444457795885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,7168,0.050789333052105375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,6144,0.04615288972854614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,1536,0.008480889101823172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,1536,0.018212444252438016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,1024,0.00685066646999783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,1024,0.015551111764378019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,5120,0.04231911235385471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,4096,0.03859377900759379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,3584,0.03789955708715651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,768,0.005963555640644497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,768,0.014862222803963555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,512,0.004307555655638377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,512,0.014335110783576965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,256,0.003576000117593341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,256,0.013527111046844058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,3072,0.035804443889194064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,128,0.0032755554550223877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,128,0.013012444807423485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,2560,0.03522133496072557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,64,0.003040888864133093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,64,0.01295555548535453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,2048,0.03358133302794562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,5120,32,0.0032702222880389956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,5120,32,0.013100444442696042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,65536,0.19416977299584282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,65536,0.2165128919813368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,16384,0.05749244160122342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,16384,0.06277689006593493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,1536,0.03317600157525804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,12288,0.038887110021379255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,12288,0.050982223616706006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,1024,0.030573334958818223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,10240,0.033731556600994535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,768,0.029656890365812514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,10240,0.045243554645114474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,512,0.0270453327231937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,256,0.025264889001846313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,8192,0.02695200012789832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,8192,0.03717511230044895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,7168,0.02436711059676276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,7168,0.03379999929004245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,6144,0.021902221772405837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,6144,0.030377778742048476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,5120,128,0.023638221952650283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,5120,0.018629334039158292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,5120,0.02754666739039951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,65536,0.24075023333231607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,4096,0.015173332558737861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,16384,0.07844355371263292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,4096,0.024010666542583044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,3584,0.013526221944226159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,3584,0.02296888828277588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,12288,0.06542933649486966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,3072,0.011990221838156382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,3072,0.021624000536070928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,10240,0.059227552678849965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,8192,0.05128711130883959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,2560,0.010491555763615502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,2560,0.020281778441535104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,2048,0.009073778159088558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,2048,0.01921688848071628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,7168,0.04793955551253425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,1536,0.007629333270920648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,6144,0.043353776137034096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,1536,0.016587555408477783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,1024,0.006216889040337668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,1024,0.015100444356600443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,5120,0.040384001202053495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,768,0.005221333354711533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,768,0.014740443891949125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,4096,0.036680890454186335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,512,0.0041244443919923575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,3584,0.03621955712636312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,512,0.01428711083200243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,256,0.003663111064169142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,256,0.013617777989970313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,3072,0.034188446071412824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,128,0.0032071111102898917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,128,0.012855999999576144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,2560,0.03421688742107815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,64,0.003128000017669466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,64,0.012987555729018318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,1536,0.03202044301562839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,2048,0.032327112224366926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,4096,32,0.0032186667538351486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,4096,32,0.012724444270133972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,65536,0.17093422677781847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,65536,0.20633422003851998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,16384,0.05109955535994636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,1024,0.028205331828859117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,16384,0.06068888637754652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,12288,0.03498133354716831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,12288,0.04945333467589485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,10240,0.03345955411593119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,10240,0.043678220775392324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,768,0.027760000692473516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,8192,0.023782221807373896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,8192,0.03560533457332187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,512,0.026850667264726426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,7168,0.021729777256647747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,256,0.024840889705551997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,7168,0.03219822380277846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,6144,0.019479999939600628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,6144,0.029206222958034937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,4096,128,0.02335022224320306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,5120,0.016311999824311998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,5120,0.025798221429189045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,16384,0.07765066623687744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,12288,0.06329600016276042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,65536,0.22515556547376844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,4096,0.013271110753218332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,4096,0.023471110396915015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,3584,0.012107555237081317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,10240,0.05774755610360039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,3584,0.022272000710169475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,3072,0.01129244433508979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,3072,0.021134222547213238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,2560,0.009908444351620143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,2560,0.019708444674809773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,8192,0.04972799950175815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,2048,0.00868977771864997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,2048,0.018776888648668926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,7168,0.046655999289618603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,1536,0.0074471111098925276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,6144,0.0423351095782386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,1536,0.01626311077011956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,1024,0.0058746664888328975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,1024,0.015007999208238391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,5120,0.03942311141226027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,768,0.0047128887640105355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,768,0.014578666951921252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,4096,0.03590044379234314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,512,0.004020444634887907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,512,0.01425155500570933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,3584,0.03567555546760559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,256,0.0034266666819651923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,3072,0.03385600116517808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,256,0.013364444176355997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,128,0.0033493333806594214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,128,0.01294222225745519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,2560,0.033815999825795494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,64,0.003101333354910215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,64,0.01292888902955585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3584,32,0.0032311110860771606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3584,32,0.01278222186697854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,2048,0.03220266766018338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,65536,0.14458666907416448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,65536,0.20215200053320992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,16384,0.04278577698601616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,16384,0.05865866608089871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,1536,0.0318195554945204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,12288,0.03487022386656867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,12288,0.047803554270002574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,1024,0.026983110441101923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,10240,0.02997066577275594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,10240,0.0421377784676022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,768,0.027730666928821143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,8192,0.021516443954573736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,8192,0.03389244609408908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,256,0.02488800055450863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,512,0.026537777649031744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,7168,0.019366222951147292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,7168,0.031593776411480375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3584,128,0.023610666394233704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,6144,0.017678222722477384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,6144,0.028559999333487615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,5120,0.016686222619480558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,5120,0.02572800053490533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,65536,0.21836177508036295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,4096,0.01333866682317522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,4096,0.023235556152131822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,16384,0.07327289051479764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,3584,0.011407111254003314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,3584,0.022014222211307947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,12288,0.06142933501137627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,3072,0.010660444696744284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,10240,0.056016888883378774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,3072,0.02076088885466258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,2560,0.00926399976015091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,2560,0.019573332534896005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,8192,0.047525333033667676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,2048,0.008044444852405125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,2048,0.018252443936136033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,7168,0.04442666636572944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,6144,0.040864888164732195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,1536,0.007090667055712805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,1536,0.015636444091796875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,1024,0.005392000079154968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,1024,0.014879110786649915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,5120,0.03839022252294753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,768,0.004707555390066571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,768,0.01441511180665758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,4096,0.035840888818105064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,512,0.004008000095685323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,512,0.014148443937301636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,3584,0.035120887888802424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,256,0.003536000019974179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,256,0.013399110900031196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,3072,0.03364888827006022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,128,0.0031226666437255014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,128,0.013053333593739403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,2560,0.033596442805396184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,64,0.0029066666546795103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,64,0.012786666552225748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,3072,32,0.003085333233078321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,3072,32,0.012720888687504662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,65536,0.12511821587880453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,65536,0.19121155473921034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,2048,0.0313937763373057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,16384,0.03835377759403653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,16384,0.05648710992601183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,1536,0.030352890491485596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,12288,0.02805688977241516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,1024,0.02719377809100681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,12288,0.04609600040647718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,10240,0.024223110742039148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,768,0.027453333139419556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,10240,0.03972355524698893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,8192,0.019814221395386588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,8192,0.03232889042960273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,512,0.02681244413057963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,7168,0.017995556195576985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,7168,0.03014755580160353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,256,0.024656888511445787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,6144,0.01573244399494595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,6144,0.02712533374627431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,3072,128,0.023544000254737005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,5120,0.013266666895813413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,5120,0.02467022173934513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,65536,0.2142417828241984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,4096,0.011074666347768573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,4096,0.0221742226017846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,16384,0.07139644357893202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,3584,0.010270222193664974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,3584,0.021394666698243883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,12288,0.059691554970211454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,3072,0.009319999979601966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,10240,0.054548445675108165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,3072,0.020057777563730877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,2560,0.008432000047630733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,2560,0.01907111042075687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,8192,0.046462221278084644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,2048,0.007473777565691207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,2048,0.01684977776474423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,7168,0.043952001465691455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,6144,0.040320889817343816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,1536,0.006491555521885554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,1536,0.015453333655993143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,1024,0.004851555658711327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,1024,0.014697778556081982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,5120,0.03834133346875509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,768,0.004262222184075249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,768,0.014392889208263822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,4096,0.03491377830505371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,512,0.0037395556767781577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,512,0.013844444519943662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,3584,0.03496799866358439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,256,0.003362666608558761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,256,0.013239111337396832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,3072,0.03308799862861633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,128,0.003075555587808291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,128,0.012706666356987424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,2560,0.033174223370022245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,64,0.0029111111329661477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,64,0.012639111114872826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,2048,0.031331555710898504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2560,32,0.0029155556112527847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2560,32,0.012683555483818054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,65536,0.10373066531287299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,65536,0.18911288844214547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,16384,0.032051556640201144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,16384,0.05428266525268555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,1536,0.030237333642111883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,12288,0.02532888948917389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,12288,0.04362755682733324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,1024,0.026597332623269822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,10240,0.022279999322361414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,10240,0.039071109559800886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,768,0.02730400032467312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,8192,0.018800000349680584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,8192,0.031614220804638334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,512,0.02679288884003957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,7168,0.01683999929163191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,7168,0.029271112547980413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,256,0.02479466630352868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,6144,0.015095111396577625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,6144,0.02701688806215922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2560,128,0.023416888382699754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,5120,0.01349422252840466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,5120,0.024193776978386775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,16384,0.0690408878856235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,4096,0.010680889089902243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,65536,0.2037057744132148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,4096,0.022198221749729578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,3584,0.009933333430025313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,3584,0.021116443806224402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,12288,0.05682933330535889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,3072,0.009089777866999308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,3072,0.01976355579164293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,10240,0.05166577630572849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,2560,0.007803555164072249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,2560,0.018582221534517076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,8192,0.044399112462997437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,2048,0.00700711127784517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,2048,0.016295111841625638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,7168,0.042066666815016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,1536,0.0058373332851462895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,1536,0.015338665909237333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,6144,0.039248890346950956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,1024,0.004511111312442356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,1024,0.014662222729788886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,5120,0.03693688909212748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,768,0.004108444270160464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,768,0.014312000738249885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,4096,0.03421333432197571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,3584,0.03432088759210374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,512,0.0037600000699361167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,512,0.013861333330472311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,256,0.0032479998966058097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,256,0.013196444345845116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,3072,0.032652444309658475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,128,0.0029893333299292457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,128,0.012916444076432122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,2560,0.03279733326699998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,64,0.002817777709828483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,64,0.012797333300113678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,2048,32,0.002893333426780171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,2048,32,0.012670222255918713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,65536,0.079839997821384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,65536,0.17938400639428032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,2048,0.02944533361328973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,16384,0.024255999260478552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,16384,0.051706665092044406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,1536,0.028741333219740126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,12288,0.020184000333150227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,12288,0.04060711132155524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,1024,0.0266977780395084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,10240,0.017913776967260573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,10240,0.03588177760442098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,768,0.027236445082558527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,8192,0.015209777487648858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,8192,0.02976800004641215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,256,0.024716445141368445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,512,0.026339555780092876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,7168,0.013991110854678683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,7168,0.028065777487225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,6144,0.01330933305952284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,6144,0.025937777426507738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,2048,128,0.023230221536424425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,5120,0.011238222320874533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,5120,0.02383377816942003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,65536,0.1998142268922594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,4096,0.009518221848540837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,4096,0.021458667185571458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,16384,0.06646133131451078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,3584,0.008931555681758458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,3584,0.020664888951513503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,12288,0.055620445145501025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,3072,0.008151111503442129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,3072,0.018442667192882962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,10240,0.050640887684292264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,2560,0.00759733302725686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,2560,0.01738577749994066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,8192,0.043812443812688194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,2048,0.006602666858169768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,2048,0.016858667135238647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,7168,0.04160711169242859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,1536,0.005457777943876054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,1536,0.015207111835479736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,6144,0.03872266742918227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,1024,0.004520888957712385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,1024,0.01460622251033783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,5120,0.03647999962170919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,768,0.004067555483844546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,768,0.014011555247836642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,4096,0.034068445364634194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,512,0.0036319999231232535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,512,0.013835555149449242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,3584,0.034400890270868935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,256,0.0032382222513357797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,256,0.013215111361609565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,3072,0.032243556446499295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,128,0.002944888960984018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,128,0.01261333293384976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,2560,0.03016177813212077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,64,0.002814222127199173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,64,0.012709333664841123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1536,32,0.0028604444944196274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1536,32,0.012535110943847232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,65536,0.05570755402247111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,65536,0.1757440037197537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,2048,0.02903466754489475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,16384,0.01856355534659492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,16384,0.04871111114819845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,1536,0.02824266751607259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,12288,0.015061333775520325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,12288,0.038704888688193426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,1024,0.02646488944689433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,10240,0.013292444248994192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,10240,0.034437331888410784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,768,0.02698933415942722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,8192,0.011382222175598145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,8192,0.029305779271655615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,512,0.02645688917901781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,7168,0.012439111040698158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,7168,0.027479110492600337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,256,0.024640000528759424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,6144,0.01035377797153261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,6144,0.025441777375009324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1536,128,0.023237332701683044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,5120,0.010120888551076254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,5120,0.02330222229162852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,65536,0.19564799467722574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,4096,0.009161777794361115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,4096,0.020019556085268658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,16384,0.0641182197464837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,3584,0.008173333273993598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,3584,0.01900622248649597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,12288,0.05221155616972181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,3072,0.007708444363541073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,3072,0.017701332767804463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,10240,0.04862666792339749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,2560,0.006588444527652528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,2560,0.016971555021074083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,8192,0.042499555481804736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,2048,0.005780444376998478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,2048,0.0159288893143336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,7168,0.040528890159395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,1536,0.004908444566859139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,1536,0.015252444479200574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,6144,0.03792533278465271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,1024,0.0042275554604000514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,1024,0.014677332507239448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,5120,0.03592355714903937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,768,0.003818666769398583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,768,0.013993778162532382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,4096,0.03233777814441257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,512,0.0034640000926123727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,512,0.013644444445768992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,3584,0.03172533379660712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,256,0.003149333306484752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,256,0.012980444563759698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,3072,0.02961155441072252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,128,0.0029360000044107437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,128,0.012643555800120035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,2048,0.028130667077170476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,64,0.002805333377586471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,64,0.012636444634861417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,2560,0.02975644336806403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,1024,32,0.0028417776856157514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,1024,32,0.0125608891248703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,65536,0.052367998494042285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,65536,0.17602666219075522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,16384,0.01498933302031623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,16384,0.0477751096089681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,1536,0.028316444820827905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,12288,0.01328000012371275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,12288,0.03860177927547031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,1024,0.02622844444380866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,10240,0.011993777420785693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,10240,0.0344471103615231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,768,0.027060444156328838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,8192,0.010806221928861407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,8192,0.029256890217463177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,512,0.02606755495071411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,7168,0.010247111320495605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,7168,0.027098667290475633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,256,0.024558222956127588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,6144,0.009679999616410997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,6144,0.025042666329277888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,1024,128,0.023031110564867657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,5120,0.009568000005351173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,5120,0.02253866692384084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,65536,0.1946115493774414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,4096,0.008639111287064022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,4096,0.019762666689025033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,16384,0.06318844689263238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,3584,0.008210666477680206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,3584,0.019104000594880845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,12288,0.052616887622409396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,3072,0.007400888535711501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,3072,0.017842666970358956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,10240,0.04878933231035868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,2560,0.006343110981914733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,2560,0.017110221915774874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,8192,0.04270755582385593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,2048,0.0054551110499435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,2048,0.015962666935390897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,7168,0.04057866666052077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,1536,0.004872888740566042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,1536,0.015245333313941956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,6144,0.037427554527918495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,1024,0.004061333421203825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,1024,0.014490667316648694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,5120,0.034986668162875705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,768,0.0037466668420367767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,768,0.0140364451540841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,4096,0.031209776798884075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,512,0.0034231110993358824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,512,0.013629333012633853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,3584,0.03122133347723219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,256,0.003069333318206999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,256,0.013173333472675748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,3072,0.029743999242782593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,128,0.0028195555011431375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,128,0.012664000193277994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,2560,0.029799110359615747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,64,0.0026933333526055017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,64,0.01290933373901579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,768,32,0.0027351110345787476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,768,32,0.012641777594884237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,65536,0.03981066743532816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,65536,0.1734257804022895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,2048,0.028121779362360638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,16384,0.012229333321253458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,16384,0.04666044314702352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,1536,0.02808711263868544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,12288,0.010747555229398938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,12288,0.038062221474117704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,1024,0.026178666286998328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,10240,0.010074666804737514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,10240,0.03370133373472426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,768,0.02682400080892775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,8192,0.0090844440791342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,8192,0.02863111098607381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,512,0.02606755495071411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,7168,0.008604444563388824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,256,0.024490666058328416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,7168,0.026568889617919922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,6144,0.00977600034740236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,6144,0.024178667200936213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,768,128,0.023042667243215773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,5120,0.009072889056470659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,5120,0.022435555855433147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,65536,0.19470400280422637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,4096,0.007819555699825287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,16384,0.06220622195137871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,4096,0.019664888580640156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,3584,0.007237333390447829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,3584,0.0188417782386144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,12288,0.05193777879079183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,3072,0.00674311113026407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,3072,0.01772799922360314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,10240,0.0483439995182885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,2560,0.006148444281684027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,2560,0.016748444901572335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,8192,0.042522665527131825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,2048,0.005466666486528184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,2048,0.015881778465376962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,7168,0.04035288757748074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,6144,0.03712266683578491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,1536,0.0047866664826869965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,1536,0.01514311134815216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,1024,0.004068444586462445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,1024,0.014352000421947904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,5120,0.03514666689766778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,768,0.0037173334923055437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,768,0.014003554979960123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,4096,0.03114933437771267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,512,0.003350222276316749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,512,0.013566222455766467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,3584,0.03121244576242235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,256,0.003060444361633725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,256,0.01316088851955202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,3072,0.029289778735902574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,128,0.002806222273243798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,128,0.012542222109105853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,2560,0.029447111818525527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,64,0.002732444347606765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,64,0.01257777793539895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,512,32,0.002936888900068071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,512,32,0.012435555458068848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,65536,0.028192000256644353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,65536,0.169649773173862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,2048,0.02791555391417609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,16384,0.01055466632048289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,16384,0.04557777775658501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,1536,0.02840533190303379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,12288,0.009341333475377824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,12288,0.037329776419533625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,1024,0.026329777306980554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,10240,0.008432889150248634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,10240,0.03331822156906128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,768,0.026838221483760413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,8192,0.007344000041484833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,8192,0.027830223242441814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,512,0.026311111119058397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,7168,0.006983111302057902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,7168,0.025461332665549383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,256,0.024364444944593642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,6144,0.006688888702127669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,6144,0.02353422178162469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,512,128,0.023038221730126276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,5120,0.006896889044178857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,5120,0.021299555897712708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,65536,0.1942755513721042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,4096,0.006520000182920032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,4096,0.019468444916937087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,16384,0.061230222384134926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,3584,0.0069724445541699724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,3584,0.01850311126973894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,12288,0.051628443929884166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,3072,0.006643555644485686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,3072,0.017658667431937326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,10240,0.04805511236190796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,2560,0.00599911105301645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,2560,0.01658933361371358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,8192,0.040234665075937905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,2048,0.005320888840489917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,2048,0.015944000747468736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,7168,0.037845333417256675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,1536,0.00462400002612008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,1536,0.015053333507643806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,6144,0.03506222367286682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,1024,0.004011555678314633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,1024,0.014530667000346713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,5120,0.03302755620744493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,768,0.00361333332127995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,768,0.013960000541475085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,4096,0.03074044320318434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,512,0.0033200000309281876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,512,0.013551111022631327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,3584,0.03072711163096958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,256,0.002973333415057924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,256,0.012889777620633444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,3072,0.029158221350775823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,128,0.0028151110228565005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,128,0.012529777983824411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,2560,0.02937777837117513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,64,0.002700444517864121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,64,0.012492444780137805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,256,32,0.0027742222365405825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,256,32,0.01241422196229299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,65536,0.017882666654056974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,65536,0.17008977466159395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,2048,0.027694221999910142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,16384,0.008060444560315874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,16384,0.04605955547756619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,1536,0.028355555401908025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,12288,0.006704000135262807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,12288,0.03657066822052002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,1024,0.025935111774338618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,10240,0.006867555694447623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,10240,0.03257955445183648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,768,0.026847999956872728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,8192,0.00675644435816341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,8192,0.027284445034133062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,512,0.026149332523345947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,7168,0.006610666712125142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,7168,0.025182222326596577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,128,0.023264888260099623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,6144,0.006490666833188798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,6144,0.023208889696333144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,256,256,0.024363555841975745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,5120,0.0068862222962909276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,5120,0.021249777740902368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,65536,0.19432711601257324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,4096,0.0064177778032090925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,4096,0.019368888603316415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,16384,0.06133333179685804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,3584,0.0068853331936730286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,3584,0.018639999959203932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,12288,0.050755557086732656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,3072,0.006545777950021956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,3072,0.017472000585661996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,10240,0.04806844393412272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,2560,0.005995555470387141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,2560,0.016721778445773654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,8192,0.039727999104393855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,2048,0.005336888962321811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,2048,0.01590666671593984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,7168,0.03790222273932563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,1536,0.004596444467703502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,1536,0.015000000596046448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,6144,0.034895999564064875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,1024,0.003968888686762916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,1024,0.014454222387737699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,5120,0.03303555647532145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,768,0.003641777982314428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,768,0.013890667094124688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,4096,0.030619555049472388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,512,0.0033173333439562055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,512,0.013367999758985309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,3584,0.030791110462612573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,256,0.0029688889367712867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,256,0.012776888906955719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,3072,0.029205332199732464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,128,0.0027439999911520216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,128,0.012479999827014076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,2560,0.029542220963372126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,64,0.0026400000270869997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,64,0.0125591109196345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,128,32,0.0026515554636716843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,128,32,0.012415111064910889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,65536,0.012862222062216865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,65536,0.16950399345821807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,16384,0.006690666493442323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,16384,0.04475288920932346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,12288,0.006568888823191325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,12288,0.03524622321128845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,10240,0.006824888702895906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,10240,0.03159911102718777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,8192,0.006697777658700943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,8192,0.027290667096773785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,7168,0.006528888725572162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,7168,0.025224000215530396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,6144,0.0063884444534778595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,6144,0.023294222023752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,5120,0.006796444455782573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,5120,0.021358221769332886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,4096,0.00637155564294921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,4096,0.01955111159218682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,3584,0.006824000014199152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,3584,0.018591110905011494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,3072,0.006509333435032103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,3072,0.017647110753589206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,2560,0.006002666635645761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,2560,0.016743999388482835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,2048,0.0053137776752312975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,2048,0.015815999772813585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,1536,0.0046657779150538975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,1536,0.015088889333936902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,1024,0.003986666599909465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,1024,0.014441778262456259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,768,0.0036435553597079385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,768,0.013948443863126965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,512,0.003294222263826264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,512,0.013487111363146039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,256,0.0029502221279674103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,256,0.01295199990272522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,128,0.0028026666906144884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,2048,0.02748888896571265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,128,0.012456888953844706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,64,0.002597333242495855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,64,0.012496888637542725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,64,32,0.002593777866827117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,64,32,0.012349333200189801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,65536,0.011696000066068439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,16384,0.006569777925809224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,65536,0.17022755410936144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,16384,0.04545155498716566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,12288,0.006447111152940326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,12288,0.035311109489864774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,10240,0.006706666615274217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,8192,0.006537777682145436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,10240,0.03168977631462945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,1536,0.02813777658674452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,8192,0.027334221535258826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,7168,0.006350222147173352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,7168,0.025248888466093276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,6144,0.006244444598754247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,5120,0.006628444625271692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,6144,0.023315555519527856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,5120,0.021167111065652635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,4096,0.006144000010357962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,4096,0.01942311061753167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,3584,0.006589333216349284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,3072,0.006246222390068903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,3584,0.01846577723821004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,3072,0.017698667115635343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,2560,0.00592533333433999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,2560,0.016772444049517315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,2048,0.0052675555149714155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,768,0.026633777552180823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,1536,0.004590222405062782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,1024,0.025892444782786902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,2048,0.015832889411184523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,1536,0.015023110641373528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,1024,0.003950222084919612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,1024,0.014473777678277759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,768,0.013859555953078799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,768,0.0035991109907627106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,512,0.003240888938307762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,256,0.0029537777105967202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,512,0.013495999905798169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,64,0.002595555658141772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,128,0.0027359999302360746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,256,0.012817777693271637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,128,0.01240888900227017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,128,32,32,0.0026062221990691293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,64,0.012493333054913415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,128,32,32,0.012351999680201212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,16384,0.3885022269354926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,12288,0.30299199952019584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,16384,0.7219608624776205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,12288,0.5513973236083984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,512,0.02574933403068119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,10240,0.2813199890984429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,10240,0.45846398671468097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,256,0.024166221419970196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,8192,0.36742220984564883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,8192,0.20471555656856963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,7168,0.3198453320397271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,7168,0.18518933984968397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,6144,0.2784293227725559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,6144,0.15840089321136475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,5120,0.23330667283799914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,5120,0.150508443514506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,128,128,128,0.02306488818592495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,8192,0.2493413289388021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,7168,0.20224621560838488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,4096,0.18916711542341444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,10240,0.3020622200436062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,4096,0.11189333597819011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,3584,0.1660426722632514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,3584,0.10007822513580322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,6144,0.1953537729051378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,3072,0.14410932858784994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,3072,0.08833689159817165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,16384,0.42270755767822266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,5120,0.16941511631011963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,2560,0.13869688245985243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,2560,0.07706666655010648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,2048,0.09876622094048394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,2048,0.06502844227684869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,12288,0.39641422695583767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,1536,0.07631466786066692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,4096,0.13745421833462185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,1536,0.05523022347026401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,1024,0.05344533258014255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,3584,0.1269128852420383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,1024,0.042393777105543345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,3072,0.10853688584433661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,2560,0.09859822193781535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,768,0.04198844565285576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,2048,0.08412266439861721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,768,0.036679112248950534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,512,0.030766222212049697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,1536,0.07903733518388537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,512,0.046224001381132335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,256,0.02387733260790507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,1024,0.05894400013817681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,512,0.03151822090148926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,768,0.05274044473965963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,256,0.017906667457686532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,128,0.010604444477293225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,256,0.037808888488345675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,128,0.02070488863521152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,64,0.007948444121413762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,65536,128,0.031932443380355835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,64,0.023874666955735948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,65536,32,0.007294221884674496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,65536,32,0.023365333676338196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,16384,0.17752444744110107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,65536,0.4248853259616428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,16384,0.11981511116027832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,12288,0.1362640062967936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,12288,0.09554133150312637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,65536,0.7356648974948459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,10240,0.11429688665601943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,10240,0.08629600207010905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,8192,0.09238221910264756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,8192,0.06694577799903022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,7168,0.08143733607398139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,7168,0.06063822242948744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,16384,0.1381440030203925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,12288,0.12276711728837754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,6144,0.07104533248477511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,10240,0.09849333100848728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,8192,0.08228444390826754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,6144,0.05545333358976576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,5120,0.05981422132915921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,5120,0.04754844307899475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,4096,0.04909600151909722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,7168,0.09220533238516913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,4096,0.04063466522428725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,3584,0.04349777764744229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,6144,0.0697831114133199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,5120,0.06377510892020331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,3584,0.042661332421832614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,3072,0.038198222716649376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,3072,0.034401777717802264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,2560,0.03342399994532267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,4096,0.056327998638153076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,3584,0.0539244446489546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,2560,0.032015111711290144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,3072,0.05009777678383721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,2048,0.027237332529491846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,2048,0.028348445892333984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,2560,0.04860088891453213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,1536,0.021540444758203294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,2048,0.04407644602987501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,1536,0.024194666081004675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,1024,0.03596355517705282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,1024,0.017237333787812125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,1024,0.02073955535888672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,768,0.011949333051840464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,1536,0.03882044553756714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,768,0.01936177743805779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,512,0.008960888617568547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,768,0.03384177883466085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,512,0.018132444885041978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,256,0.00630666646692488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,128,0.013756444056828817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,512,0.03164266546567281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,256,0.01482488877243466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,128,0.0041759999261962045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,65536,0.4779937532212999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,32,0.004179555508825514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,256,0.026138666603300307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,16384,64,0.0038693332009845306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,16384,128,0.024409777588314478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,64,0.014465777410401238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,16384,32,0.014464888307783337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,16384,0.14784088399675158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,65536,0.5400062137179905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,16384,0.10467200146781074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,65536,0.38366932339138454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,12288,0.10570044649971856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,12288,0.0829191141658359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,10240,0.09029066562652588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,10240,0.07024977604548137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,8192,0.07201600074768066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,8192,0.05899288919236925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,7168,0.06312800116009183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,12288,0.0974951121542189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,16384,0.11833688947889541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,7168,0.05339822173118591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,10240,0.08692444695366754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,6144,0.056045333544413246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,5120,0.04733511143260532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,5120,0.04282044370969137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,8192,0.07808711131413777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,6144,0.04743822084532844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,7168,0.07485955291324191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,6144,0.063009778658549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,4096,0.037610666619406805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,5120,0.05793066819508871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,4096,0.036130666732788086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,3584,0.033414221472210355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,3584,0.03347466720475091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,4096,0.050583998362223305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,3072,0.029822223716311987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,3072,0.030373334884643555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,3584,0.04989511105749342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,65536,0.4272933271196153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,2560,0.025425778494940862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,2560,0.028696000576019287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,3072,0.04618755645222134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,2048,0.021240888370407954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,2560,0.04488977789878845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,2048,0.02529866662290361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,2048,0.04049422343571981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,1536,0.016977777083714802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,1536,0.021894221504529316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,1024,0.011798222031858234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,1536,0.0369395547442966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,768,0.033196445968416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,1024,0.019738665885395475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,768,0.009752888646390703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,1024,0.0343537794219123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,768,0.018903111418088276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,512,0.007743111087216272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,512,0.017070222232076857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,256,0.0052684446175893145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,512,0.031007111072540283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,256,0.014491554763582019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,128,0.024117334021462336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,128,0.0038168889780839286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,12288,256,0.02554133368863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,128,0.013544888959990608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,64,0.003536888708670934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,12288,32,0.0036551112102137674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,16384,0.1321386628680759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,64,0.014362666341993542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,12288,32,0.014120888378885059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,65536,0.4722791247897678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,65536,0.3381697866651747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,16384,0.09799733426835801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,12288,0.08771555291281806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,12288,0.07797688908047147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,16384,0.10781066947513157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,10240,0.07356711228688557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,10240,0.07096799876954821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,12288,0.10467733277214898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,8192,0.059248891141679555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,8192,0.05502222312821282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,7168,0.052601777844958834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,10240,0.0787386695543925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,7168,0.049694223536385425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,6144,0.04674933354059855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,5120,0.03950488898489211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,8192,0.06688355737262301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,6144,0.044920888211992055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,7168,0.07200355662239923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,5120,0.03994755612479316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,4096,0.032495998673968844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,65536,0.3817884392208523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,3584,0.03157777918709649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,4096,0.034861332840389676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,6144,0.06273688872655232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,3584,0.029727998707029555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,5120,0.054379555914137095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,4096,0.04826044374042087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,3072,0.02572888963752323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,3584,0.04732444551255968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,2048,0.018532445033391316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,3072,0.02906755606333415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,1536,0.014530667000346713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,2560,0.022089777721299067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,3072,0.04378133349948459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,2560,0.02716800073782603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,2560,0.04168711105982462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,2048,0.023972445064120825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,2048,0.038329776790406965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,1536,0.020921778347757127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,1024,0.010580444501505958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,512,0.007170666423108842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,1536,0.035882665051354304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,1024,0.019083556201722886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,768,0.008924444516499838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,512,0.029655999607510034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,1024,0.03388266762097677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,768,0.0181013329161538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,768,0.032607999112870954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,512,0.015791111522250705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,256,0.004237333519591225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,128,0.023734221855799358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,64,0.013991110854678683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,256,0.014399111270904541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,128,0.0037360000941488477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,10240,256,0.025584000680181716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,128,0.01367022262679206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,64,0.0034808889031410217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,10240,32,0.003581333491537306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,10240,32,0.013827555709415011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,65536,0.38704533047146267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,65536,0.2481679916381836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,16384,0.10048088762495254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,16384,0.075564444065094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,12288,0.07037422392103407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,12288,0.060384001996782094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,10240,0.05911111169391208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,16384,0.0945520003636678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,10240,0.05201422174771627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,12288,0.08449421988593207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,7168,0.04263110955556234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,65536,0.3167102336883545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,8192,0.04787733488612705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,8192,0.059398220645056836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,8192,0.044347554445266724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,10240,0.07882933484183417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,7168,0.04053777787420485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,6144,0.03775022096104092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,6144,0.03696533375316196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,7168,0.05628177854749891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,5120,0.031960888041390315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,5120,0.03329777717590332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,6144,0.053094221485985645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,4096,0.026368000441127356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,4096,0.029143999020258587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,5120,0.05141333407825894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,3584,0.02348888913790385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,3584,0.026925333672099646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,4096,0.044066667556762695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,3072,0.02100177771515316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,3584,0.04244266615973579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,3072,0.024887111451890733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,2560,0.01789866718981001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,2048,0.02107200026512146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,3072,0.039956443839603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,1536,0.011617778076065911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,2560,0.0229795558585061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,2048,0.01462933255566491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,2560,0.039417776796552874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,2048,0.03607022100024753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,1536,0.01938488913906945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,1024,0.008736888567606607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,1024,0.01791733337773217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,1536,0.03387377659479777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,768,0.007391999993059371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,1024,0.03286755416128371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,768,0.01628800067636702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,512,0.006020444548792309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,768,0.031930665175120033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,512,0.014483556151390076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,256,0.004031110968854693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,512,0.029375112719006006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,256,0.013474666410022311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,128,0.023757333556811016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,64,0.01311555587583118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,128,0.003527111063400904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,128,0.013011555704805585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,8192,256,0.024845333562956914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,64,0.003313777761326896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,65536,0.23099732398986816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,16384,0.08948444657855564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,8192,32,0.003641777982314428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,8192,32,0.01295733369059033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,12288,0.05757511324352688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,65536,0.3350977897644043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,16384,0.07035199801127116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,12288,0.06228533056047228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,10240,0.05265688896179199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,10240,0.04854311214552986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,12288,0.08415644698672825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,8192,0.0425528883934021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,7168,0.037789334853490196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,16384,0.09029689100053574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,8192,0.041506667931874595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,10240,0.07022400034798516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,7168,0.04068355427847968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,6144,0.03389066788885329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,7168,0.05379466546906365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,8192,0.05736088752746582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,6144,0.035008000002966985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,5120,0.028515554136700098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,5120,0.04687111245261299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,5120,0.03173689047495524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,6144,0.05067377620273166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,4096,0.023718222975730896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,4096,0.027422222826215956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,3584,0.021152888735135395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,65536,0.2899742126464844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,3584,0.025489777326583862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,4096,0.042043556769688926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,3072,0.01872711049185859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,3072,0.023372444841596816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,3584,0.04088622331619263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,2560,0.01608533329433865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,3072,0.037865777810414634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,2560,0.021551110678248938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,1536,0.01888533267709944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,2048,0.013139555851618448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,1024,0.008189332981904348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,2560,0.037824889024098717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,2048,0.02014400064945221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,1536,0.010604444477293225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,1024,0.03319733341534933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,2048,0.035418666071361966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,1024,0.01771111124091678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,512,0.014260444376203748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,1536,0.033561779393090144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,768,0.007035555938879649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,768,0.015088889333936902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,512,0.0058168888919883305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,768,0.03071199854214986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,256,0.0038151111867692736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,256,0.013392888837390475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,512,0.027090667022599116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,128,0.003378666730390655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,256,0.024991111622916326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,128,0.012895999683274163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,64,0.0033857776886887024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,64,0.013193777865833707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,7168,128,0.02388444377316369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,7168,32,0.0034648889882696998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,7168,32,0.013017777767446307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,65536,0.27959733539157444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,65536,0.21303644445207384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,16384,0.08081333504782783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,16384,0.06474577718310885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,12288,0.05767377879884508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,12288,0.05448711249563429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,10240,0.04602400130695767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,10240,0.04536088969972399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,16384,0.09904444217681885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,8192,0.037179556157853864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,12288,0.07620710796780057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,10240,0.06730755832460192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,8192,0.038896001047558255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,7168,0.033358222908443876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,7168,0.03613333238495721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,7168,0.05458577805095249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,8192,0.059439109431372754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,6144,0.029896888468000624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,65536,0.26950756708780926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,6144,0.03284710976812575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,5120,0.025359110699759588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,5120,0.029197331931855943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,4096,0.02749155627356635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,5120,0.0481697784529792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,6144,0.0511795547273424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,4096,0.02092444399992625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,3584,0.01866133345497979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,3584,0.024139554964171514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,4096,0.04129599862628513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,3072,0.017443555924627516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,3584,0.03987110985649957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,3072,0.02223288847340478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,2560,0.01404800017674764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,2560,0.02103999919361538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,2048,0.034726222356160484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,3072,0.03721333212322659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,1024,0.007599111232492659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,2048,0.011863999896579318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,2560,0.037195556693606906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,2048,0.020205333828926086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,1536,0.009720889230569204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,1536,0.018550222118695576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,1024,0.016051555673281353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,1536,0.03243733445803324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,768,0.006719111154476802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,1024,0.03193688723776076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,768,0.014835556348164877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,512,0.004582222137186262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,768,0.02885155545340644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,512,0.014427555931939019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,256,0.003779555360476176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,512,0.026722666290071275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,128,0.023482667075263128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,256,0.013552889227867126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,128,0.0033448889023727844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,32,0.012870222330093384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,6144,256,0.024757333927684363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,128,0.012847111456924014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,64,0.0032551110618644287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,65536,0.2334222263760037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,6144,64,0.012947555217478009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,6144,32,0.003318222239613533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,65536,0.20042488310072157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,16384,0.06789688931571113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,16384,0.06100177764892578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,10240,0.03942133320702447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,12288,0.053852445549435086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,12288,0.05026488833957248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,10240,0.042918221818076245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,8192,0.036701334847344294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,16384,0.07945066690444946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,8192,0.031896889209747314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,12288,0.06646133131451078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,7168,0.028745777077145044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,10240,0.05897777610354953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,7168,0.03427911135885451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,8192,0.05131466521157158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,6144,0.02578488820128971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,6144,0.03102311160829332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,7168,0.049273778994878135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,5120,0.021884444687101576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,5120,0.027830223242441814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,6144,0.045921779341167875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,4096,0.01791733337773217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,4096,0.024318221542570326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,5120,0.04292444388071696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,3584,0.016255999604860943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,4096,0.03791911072201199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,3584,0.023017777336968318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,3072,0.01419911119672987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,3584,0.037412444750467934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,65536,0.36559377776251895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,3072,0.021557332740889654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,2560,0.012168000141779581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,3072,0.03538489010598924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,2560,0.02069244450993008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,1536,0.017999110950364005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,2048,0.010216000179449717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,2048,0.019703111714786954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,2560,0.03530399998029073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,1536,0.00850755555762185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,2048,0.03788977861404419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,1024,0.006808888994985157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,1536,0.03202755583657159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,1024,0.01552177800072564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,768,0.005886222339338726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,1024,0.031919111808141075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,256,0.003592888928121991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,768,0.014781332678265043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,512,0.004505777938498391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,768,0.029061334000693426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,512,0.014391111003028022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,128,0.013007111019558378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,512,0.026309332913822595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,256,0.013409777647919126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,128,0.0032275555034478507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,256,0.024586665961477492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,64,0.0030382221771611106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,5120,128,0.023541332946883306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,64,0.013031110995345645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,5120,32,0.0031439999325407874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,65536,0.19462754991319445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,16384,0.05227377679612902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,5120,32,0.012791111237472959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,65536,0.188408891359965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,16384,0.05910044246249729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,12288,0.04267644551065233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,12288,0.051515556044048734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,10240,0.03692622317208184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,16384,0.073761781056722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,10240,0.041103111373053655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,8192,0.03537155522240533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,12288,0.06226933002471924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,8192,0.031424889961878456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,7168,0.02823466724819607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,7168,0.03273066547181871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,10240,0.057908442285325795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,8192,0.048375109831492104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,6144,0.02566577825281355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,65536,0.2433075639936659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,6144,0.029883556895785864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,7168,0.046331554651260376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,5120,0.01829866733815935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,6144,0.04380711250834995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,5120,0.026723555392689172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,4096,0.014953777194023132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,5120,0.04134044382307265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,4096,0.024285333024130926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,3072,0.021384000778198242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,3584,0.013339555925793119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,4096,0.037540445725123085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,3584,0.02270222206910451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,3072,0.011754666765530905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,3584,0.03653511073854234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,2048,0.019183110859658983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,2560,0.03469244307941861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,2560,0.010424888796276517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,2048,0.032983111010657415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,2560,0.020284444093704224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,3072,0.03501866592301263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,1024,0.015370666980743408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,2048,0.008999110923873054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,1536,0.007634667058785756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,1536,0.01664533383316464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,1024,0.006180444525347815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,1536,0.0314284430609809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,768,0.005190222213665645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,256,0.003370666669474708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,1024,0.028446220689349707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,768,0.014871999621391296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,512,0.0037431112594074676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,768,0.028010666370391846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,512,0.014367111855083041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,512,0.026018665896521673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,256,0.013463111387358772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,128,0.003108444520168834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,128,0.012853333519564735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,32,0.012639111114872826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,256,0.02441155579355028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,64,0.0029688889367712867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,4096,128,0.023128888673252527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,4096,64,0.012823111481136747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,16384,0.0565777752134535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,4096,32,0.0031519999934567344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,65536,0.1744684378306071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,12288,0.04677599999639723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,16384,0.047482666042115956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,65536,0.18365511629316542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,10240,0.03989244500796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,12288,0.06138133340411716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,12288,0.03861155443721347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,16384,0.07105066378911336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,10240,0.03326933251486884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,7168,0.022985777921146814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,8192,0.028100444210900202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,10240,0.0549777779314253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,8192,0.035416887866126165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,8192,0.046910223033693105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,65536,0.22919376691182455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,7168,0.031098667118284438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,6144,0.020356444848908316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,6144,0.028132445282406274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,7168,0.04550844430923462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,5120,0.016568000117937725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,6144,0.04322310951020983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,5120,0.02513511147763994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,4096,0.013281777501106262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,3584,0.021947556071811255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,5120,0.039820445908440485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,4096,0.022943110929595098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,3584,0.012086222569147745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,4096,0.03640266590648227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,3072,0.01093777765830358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,2560,0.019746666153271992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,3584,0.0361422234111362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,3072,0.020863110820452373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,2048,0.018760000665982563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,2560,0.00963199966483646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,3072,0.03391199972894456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,2048,0.008268444074524773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,2560,0.034662223524517484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,1536,0.007207999626795451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,2048,0.03292977809906006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,1536,0.016044444508022733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,1024,0.00600888869828648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,768,0.01459377838505639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,1536,0.030838221311569214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,1024,0.015216888652907478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,768,0.027439999911520217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,768,0.004664888812435998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,1024,0.028210666444566514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,256,0.013428444663683573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,512,0.00371111101574368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,512,0.014055111342006259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,256,0.0033119999700122406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,512,0.02594311038653056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,128,0.00309244439833694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,256,0.02459200057718489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,32,0.012602667013804117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,128,0.012866666747464074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,64,0.0029208889851967492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3584,128,0.022991110881169636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,16384,0.04574222366015116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3584,64,0.012713777522246042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3584,32,0.00305244450767835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,65536,0.14566132757398817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,65536,0.17985244592030844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,16384,0.05475644601715935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,12288,0.0341360006067488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,12288,0.044807109567854136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,16384,0.07110666566424899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,10240,0.029304001066419814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,12288,0.060862223307291664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,10240,0.03863911165131463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,8192,0.025024889243973628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,8192,0.03282489048110114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,10240,0.05386933353212145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,65536,0.2263422277238634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,7168,0.024337778488794964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,8192,0.04554488923814562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,7168,0.030588444736268785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,6144,0.017266665895779926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,6144,0.027642667293548584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,7168,0.04822399881150988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,4096,0.022653333014912073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,5120,0.03957155678007338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,5120,0.0143857780430052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,4096,0.03592711024814182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,6144,0.04224977890650431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,5120,0.02494666642612881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,4096,0.011893333660231696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,3584,0.011160000330872007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,3584,0.02164177762137519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,3072,0.010063110954231685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,3584,0.03560355636808608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,3072,0.02053600052992503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,2048,0.01852799952030182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,2560,0.00903200027015474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,3072,0.033711112207836576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,1536,0.016004444824324716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,2560,0.019592000378502738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,2048,0.00778311077091429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,2560,0.03415111038419936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,1536,0.006786666810512543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,2048,0.03267111049758063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,1024,0.005579555614127054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,1024,0.015018666783968607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,1536,0.028304888142479792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,768,0.004550222307443619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,1024,0.02791733211941189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,768,0.014702222413486905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,512,0.0037235555549462638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,768,0.027496889233589172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,512,0.014020444618331062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,256,0.0033173333439562055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,512,0.02607111136118571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,256,0.01329955541425281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,128,0.003134222287270758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,256,0.02425333360830943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,128,0.012881778180599213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,64,0.0029271110478374693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,3072,128,0.023192889160580103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,64,0.012705778082211813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,3072,32,0.002985777747299936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,3072,32,0.012729778057999082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,65536,0.1221804486380683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,65536,0.16753333144717744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,16384,0.03599822190072801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,16384,0.052489777406056724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,12288,0.029161777761247423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,12288,0.0414613319767846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,16384,0.07054133547676934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,10240,0.025553777813911438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,12288,0.059848891364203564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,10240,0.035840888818105064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,65536,0.21403200096554229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,8192,0.021021333005693223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,10240,0.053787555959489614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,8192,0.031114667654037476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,7168,0.016984888248973422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,8192,0.045724444919162326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,7168,0.02921066681543986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,6144,0.017275555266274344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,7168,0.04335822330580818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,6144,0.026854222019513447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,5120,0.015039111177126566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,6144,0.042338665988710195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,3584,0.010527999864684211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,5120,0.023744889431529578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,4096,0.010746666954623329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,4096,0.035939554373423256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,5120,0.039613333013322614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,2560,0.008367111285527548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,3072,0.019816888703240287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,4096,0.021792888641357422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,3584,0.020970667401949566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,3072,0.010545777777830759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,2048,0.01641688909795549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,3584,0.03559022148450216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,2560,0.018785778019163344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,2048,0.007238222493065729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,3072,0.033739556868871055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,2560,0.033851553996404014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,1536,0.0064328888224230866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,2048,0.031202667289310034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,1536,0.015561777684423657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,768,0.014376888672510782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,1024,0.005146666533417172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,1536,0.028428445259730022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,1024,0.014800000521871777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,256,0.00328533330725299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,768,0.004321777572234471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,1024,0.02795999911096361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,512,0.0036924444139003754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,768,0.027136888768937852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,512,0.013936888840463428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,512,0.025851555996470984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,256,0.013256000147925483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,128,0.0030657777355776895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,256,0.02439466615517934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,128,0.012873777912722694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,64,0.002895999906791581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2560,128,0.023015111684799194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,64,0.012714666624863943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2560,32,0.002934222213096089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2560,32,0.012455110748608908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,65536,0.10212177700466579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,65536,0.1643679936726888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,16384,0.06591910786098905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,16384,0.030419554975297716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,10240,0.021002666817771062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,16384,0.05004355642530653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,12288,0.0247013337082333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,12288,0.040291554398006864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,10240,0.03462488783730401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,65536,0.20232889387342665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,8192,0.017661333084106445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,12288,0.05787999762429131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,8192,0.030351112286249798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,10240,0.050901333491007485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,7168,0.04214222232500712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,7168,0.01591022147072686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,7168,0.02831022275818719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,8192,0.043192889955308705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,6144,0.014680888917711047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,6144,0.026483555634816486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,5120,0.012776888906955719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,5120,0.02370044423474206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,6144,0.040582223070992365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,4096,0.010618666807810465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,5120,0.03801777627733018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,4096,0.021780444516075984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,3584,0.009496000077989366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,4096,0.03482044405407376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,3584,0.019845333364274766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,3072,0.008983111215962304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,3584,0.034691555632485285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,3072,0.019651555352740817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,2560,0.008114666574531132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,3072,0.03289066751797994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,1536,0.006068444500366847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,2560,0.01758311192194621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,2048,0.007227555745177799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,2560,0.0314720008108351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,2048,0.01626755628320906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,2048,0.029434667693244085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,1536,0.015584000282817416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,1024,0.004469333423508538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,1536,0.02830577890078227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,1024,0.014754666222466363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,768,0.0042133331298828125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,1024,0.027857777145173814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,768,0.014282666974597506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,512,0.003592888928121991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,768,0.027424888478385076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,512,0.013807111316257052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,256,0.0031866667171319327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,512,0.02569599946339925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,256,0.013267555170589022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,128,0.0029324444217814338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,256,0.024254222710927326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,128,0.012633778154850006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,64,0.0028231110837724474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,2048,128,0.023000889354281958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,64,0.0127368892232577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,2048,32,0.002983999955985281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,2048,32,0.012588444683286877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,65536,0.08036355839835273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,65536,0.15265511141883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,16384,0.02500711050298479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,16384,0.04709511001904806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,65536,0.1821813318464491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,12288,0.053618666198518544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,12288,0.019899555378490023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,16384,0.06208799944983589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,8192,0.028540445698632136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,12288,0.03743199838532342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,10240,0.01742044422361586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,7168,0.02661422226164076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,10240,0.03247200118170844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,8192,0.01515111161602868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,10240,0.04717600014474657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,7168,0.013819555441538492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,8192,0.040912889771991305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,6144,0.012353777885437012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,7168,0.04005333450105455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,6144,0.02495288848876953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,5120,0.010998222563001843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,3584,0.008830221990744272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,6144,0.03870133227772183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,5120,0.02260977857642704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,4096,0.009495110975371467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,5120,0.036891556448406644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,4096,0.020563556088341605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,4096,0.03417155477735732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,3584,0.01923733287387424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,2560,0.01699733402993944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,3072,0.008097777764002482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,3584,0.033933334880405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,3072,0.017872000734011333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,2560,0.0074035558435651995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,3072,0.03165244393878513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,2048,0.0064479998416370815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,2048,0.01607466737429301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,2560,0.031026668018764917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,1536,0.005294222384691238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,2048,0.029230223761664495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,1536,0.015384889311260648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,1536,0.028008000718222723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,1024,0.004421333471934001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,1024,0.014730667074521383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,768,0.004051555362012652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,1024,0.027684445182482403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,768,0.01426577733622657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,256,0.013144888811641268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,512,0.0036231109665499795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,768,0.02678844498263465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,512,0.013839110732078552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,256,0.003185777821474605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,512,0.02554133368863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,32,0.0028702221396896574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,128,0.0029333333174387612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,128,0.012640888492266336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,256,0.024262222978803847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1536,64,0.002784888984428512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1536,128,0.02286311156219906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,64,0.012702222499582501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1536,32,0.012462221913867526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,65536,0.06066933605406019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,16384,0.06045866674847073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,16384,0.01809511085351308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,65536,0.14920711517333984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,16384,0.04513066675927904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,12288,0.014760888285107084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,65536,0.20805244975619844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,10240,0.048166222042507596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,12288,0.036321779092152916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,10240,0.01402844488620758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,12288,0.05375644564628601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,10240,0.031312889522976346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,8192,0.012730666332774691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,8192,0.028039998478359644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,7168,0.012051555845472546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,8192,0.04169688953293694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,7168,0.026126222478018865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,6144,0.011220444407727985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,7168,0.04024799995952182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,4096,0.009402666654851701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,6144,0.02424000038041009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,6144,0.038986666334999934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,5120,0.010445333189434476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,5120,0.022506665852334764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,5120,0.03687644335958693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,4096,0.01943111088540819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,3584,0.008504889077610439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,4096,0.03348355491956075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,3584,0.018696889281272888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,3072,0.007840000092983246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,3584,0.031629333893458046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,3072,0.017608889275126986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,2560,0.006552000012662675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,3072,0.030117332935333252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,2560,0.017157332764731515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,2048,0.005935110979610019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,1024,0.004085333396991094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,1536,0.027959111664030287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,2560,0.03049955434269375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,2048,0.01612622208065457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,1536,0.004819555415047539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,2048,0.029208001163270738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,1536,0.015461333923869662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,1024,0.014745778507656522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,768,0.0037715555065208008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,1024,0.027521777484152053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,768,0.014309333430396186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,512,0.003378666730390655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,768,0.02716711163520813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,512,0.013770666387346057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,128,0.012668444050682915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,512,0.025369778275489807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,256,0.003104000041882197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,256,0.012913777596420713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,128,0.0028942223224374983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,256,0.024119999673631456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,64,0.0027111110587914786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,1024,128,0.022898667388492163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,64,0.012572444147533841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,1024,32,0.0027253333893087176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,1024,32,0.0124177775449223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,65536,0.04741866721047295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,65536,0.14791377385457358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,16384,0.01738666660255856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,16384,0.0433555543422699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,65536,0.16815822654300264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,12288,0.014458666245142618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,16384,0.056426664193471275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,12288,0.0364951127105289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,10240,0.013054221868515015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,12288,0.05115644468201531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,10240,0.03152799937460158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,8192,0.011562666959232755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,10240,0.04548355605867174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,8192,0.027318222655190363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,7168,0.010979555547237396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,8192,0.039707554711235896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,7168,0.025661332739724055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,6144,0.010182222558392419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,7168,0.038895110289255776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,5120,0.03539911243650649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,6144,0.02366399930583106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,5120,0.0094373333785269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,6144,0.037296000454160906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,5120,0.021776888105604384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,4096,0.008200888832410177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,4096,0.01928622192806668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,3584,0.007826666865083907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,4096,0.03101244568824768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,3072,0.029742221037546795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,2560,0.01682400041156345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,3584,0.018740443719757926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,3072,0.007339555356237624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,3584,0.031125333574083116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,1536,0.004795555439260271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,3072,0.017754667335086398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,2560,0.006334222025341458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,2048,0.005498666730191972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,2560,0.030697776211632624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,2048,0.01608088943693373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,2048,0.02910311023394267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,1536,0.015436444017622205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,1024,0.00408266650305854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,1536,0.027446221974160936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,1024,0.014771555860837301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,512,0.013725333743625216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,768,0.0037786666717794207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,1024,0.02737599942419264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,768,0.014197332991494073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,128,0.00286311118139161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,512,0.00342399999499321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,768,0.026897778113683064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,256,0.0030906666070222855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,256,0.01310222182008955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,512,0.02549422283967336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,256,0.024033778243594702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,128,0.01277955538696713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,65536,0.03708444370163812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,768,128,0.02286400066481696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,64,0.0026808888134029177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,64,0.012642666697502136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,768,32,0.0027226667023367355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,768,32,0.012435555458068848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,65536,0.14482666386498347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,16384,0.013575110998418597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,16384,0.04284088810284933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,65536,0.16085778342352972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,12288,0.011687111523416309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,12288,0.03532444437344869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,16384,0.05569066603978475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,10240,0.0107342220014996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,10240,0.031136890252431233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,7168,0.00925422211488088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,12288,0.04962755574120415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,8192,0.00978577799267239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,10240,0.044403556320402354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,8192,0.027053332991070215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,8192,0.03957510987917582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,7168,0.02500088844034407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,5120,0.02068444424205356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,6144,0.008737777670224508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,6144,0.022809776994917128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,7168,0.03801511062516107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,5120,0.007671111159854465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,6144,0.03510399990611606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,4096,0.00700800038046307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,4096,0.01899377836121453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,5120,0.03289866778585646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,3584,0.007069333559936947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,4096,0.030348443322711523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,3072,0.029126223590638902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,3584,0.018407111366589863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,3072,0.006736889067623351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,3072,0.017484444710943434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,2048,0.01603200038274129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,3584,0.030637333790461223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,2560,0.006156444549560547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,1536,0.015303999185562134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,2560,0.01683022247420417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,2048,0.005432888865470886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,2560,0.02937155630853441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,1536,0.004763555609517627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,1024,0.02718844513098399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,2048,0.028781334559122723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,1024,0.004062222109900581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,1536,0.027349332968393963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,1024,0.014634667171372307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,768,0.0037093332244290244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,768,0.014020444618331062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,512,0.0036880001425743103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,512,0.013589333328935834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,768,0.026836444934209187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,512,0.02516355613867442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,64,0.0026942222482628296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,256,0.003048000029391713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,256,0.012818666795889536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,128,0.0028328889360030494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,256,0.0240657776594162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,128,0.012709333664841123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,512,128,0.022839999861187402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,512,32,0.0027093332674768236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,64,0.012485333614879184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,512,32,0.012196444802814059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,12288,0.00888000014755461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,65536,0.02397777802414364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,65536,0.14188267125023737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,16384,0.00998133338159985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,16384,0.04220088985231188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,65536,0.16084267033471003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,16384,0.0552151103814443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,12288,0.03455377618471781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,10240,0.007965332931942409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,12288,0.048384000857671104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,10240,0.029891557163662378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,8192,0.007262222468852997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,10240,0.043472889396879405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,8192,0.025506666964954797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,7168,0.0070266665683852295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,8192,0.038161777787738375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,7168,0.02379377848572201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,6144,0.0064995553758409284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,6144,0.022152889106008742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,7168,0.03577155537075467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,5120,0.006966222491529252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,6144,0.034965333011415266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,5120,0.020367999871571858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,4096,0.0064995553758409284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,5120,0.032636443773905434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,3072,0.01719555589887831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,4096,0.01871466636657715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,3584,0.0069191112286514705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,3584,0.01809866726398468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,4096,0.030045333835813735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,3072,0.006605333338181178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,3584,0.03053599927160475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,2560,0.005974222388532426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,3072,0.029120001527998183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,2560,0.016439111696349252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,2048,0.005316444569163852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,2560,0.029602666695912678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,2048,0.015684444043371413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,1536,0.0046640001237392426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,2048,0.028640889459186133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,768,0.00361333332127995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,1536,0.015048000547620984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,1024,0.00397155558069547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,1536,0.02734222180313534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,1024,0.014168888330459595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,1024,0.02734400000837114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,768,0.014001778430408902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,512,0.0032986667421129015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,768,0.02643377747800615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,512,0.01350311107105679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,256,0.002946666752298673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,512,0.025026667449209426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,256,0.012792888614866467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,128,0.0027911110470692315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,128,0.012456888953844706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,256,0.024079110887315538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,64,0.002623999905255106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,256,128,0.022685334086418152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,16384,0.0075528886583116316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,64,0.01240888900227017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,256,32,0.0026417778184016547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,65536,0.16161866982777914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,256,32,0.012156444291273752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,65536,0.014879110786649915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,65536,0.14158844947814941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,12288,0.006888888776302338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,16384,0.04176977939075894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,8192,0.006761777732107375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,10240,0.042324443658192955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,16384,0.05483822358979119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,12288,0.0336151123046875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,8192,0.0364951127105289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,7168,0.02410488824049632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,10240,0.006826666494210561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,6144,0.006396444307433234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,12288,0.046132445335388184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,10240,0.029221332735485498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,8192,0.02532711128393809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,7168,0.006582222051090664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,4096,0.0063991112013657885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,6144,0.022095110681321885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,7168,0.03570488757557339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,5120,0.006778666542636023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,6144,0.03424711028734843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,5120,0.020431111256281536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,5120,0.03281155559751723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,4096,0.018637332651350234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,3584,0.006862222320503659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,4096,0.02997066577275594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,3584,0.017903111047214933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,3072,0.006509333435032103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,3584,0.0303422212600708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,3072,0.017158221867349412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,2048,0.015555555621782938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,2560,0.005954666684071223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,3072,0.02904799911710951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,2560,0.016357334123717416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,2048,0.0052986666560173035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,2560,0.02909244431389703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,1536,0.004595555365085602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,2048,0.02852977646721734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,1536,0.014967110421922473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,1024,0.003997333347797394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,1536,0.0270506673389011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,1024,0.01425244410832723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,768,0.003590222034189436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,512,0.025145777397685583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,256,0.0029164445069101122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,1024,0.02702933384312524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,768,0.013897778259383308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,512,0.003289777785539627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,768,0.026437333888477747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,512,0.013407111167907715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,256,0.012795555922720166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,128,0.0027804445061418745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,256,0.023863111933072407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,128,0.01240888900227017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,64,0.00259911103381051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,96,128,128,0.022680888573328655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,64,0.012303999728626676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,128,32,0.0026337777574857077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,128,32,0.012238221863905588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,65536,0.01257599973016315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,16384,0.006601777755551868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,65536,0.1410177813635932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,16384,0.04023022121853299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,8192,0.02534755567709605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,12288,0.006486222147941589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,12288,0.03302133414480422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,10240,0.006753777878151999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,10240,0.028858668274349634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,8192,0.006618666566080517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,7168,0.006497777584526274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,7168,0.02368533280160692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,6144,0.006335111127959357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,6144,0.021918222308158875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,5120,0.006736889067623351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,5120,0.020304888486862183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,4096,0.0063271112740039825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,4096,0.01866044435236189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,3584,0.006769777586062749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,3584,0.017978666557206046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,3072,0.006412444429265128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,3072,0.017120889491505094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,2560,0.005938666562239329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,2560,0.0163128889269299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,2048,0.005256888767083486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,2048,0.015551111764378019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,1536,0.004600888739029567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,1536,0.014886221951908536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,1024,0.003952888978852166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,1024,0.014167111780908374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,768,0.003572444534964032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,768,0.01383733335468504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,512,0.00327822214199437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,512,0.013363555901580386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,256,0.0029235554652081597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,128,0.0027066665805048416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,256,0.012852444416946836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,128,0.012431999875439538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,64,0.0026008888251251644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,64,0.012230222423871359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,64,32,0.002591111179855135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,64,32,0.012177777787049612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,65536,0.011379555695586734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,16384,0.00654755574133661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,65536,0.1420604387919108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,16384,0.03964266512129042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,12288,0.0063991112013657885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,10240,0.006569777925809224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,12288,0.032912890116373696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,10240,0.028974221812354192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,8192,0.0064399999876817065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,8192,0.025440888272391424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,7168,0.0063635553750726916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,7168,0.023583999938435022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,6144,0.00628977765639623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,6144,0.021910222040282354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,5120,0.006613333192136552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,5120,0.020420443680551317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,4096,0.006207110981146495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,4096,0.01869244376818339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,3584,0.0064995553758409284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,3584,0.01789066692193349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,3072,0.006233777850866318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,3072,0.017080000705189176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,1024,0.003915555361244413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,2560,0.0059262220230367445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,2560,0.016343111793200176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,2048,0.005255110975768831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,768,0.013759999639458127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,512,0.003265777809752358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,2048,0.015641777051819693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,1536,0.004602666530344221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,1536,0.014882667197121514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,768,0.003567999849716822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,1024,0.014200000299347771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,256,0.002918222298224767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,64,0.012320888539155325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,512,0.013265777793195514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,256,0.012738666600651212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,128,0.0027475555737813315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,128,0.012355555262830524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,64,0.0025884444928831523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,96,32,32,0.00259911103381051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,96,32,32,0.012167111039161682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,16384,0.39568177858988446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,12288,0.3027431170145671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,16384,0.7403484450446235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,12288,0.5697973569234213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,10240,0.255383120642768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,10240,0.47406758202446836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,8192,0.22843376795450845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,8192,0.37845600975884336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,7168,0.3341573344336615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,7168,0.18386933538648817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,6144,0.30614667468600804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,6144,0.17640799946255156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,5120,0.24072710673014322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,5120,0.13574222723642984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,4096,0.19438489278157553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,4096,0.11647378073798285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,8192,0.2453706794314914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,7168,0.2160951164033678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,6144,0.1943377786212497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,3584,0.10006755590438843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,3584,0.1717217763264974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,5120,0.16785421636369494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,12288,0.35276177194383407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,10240,0.30548710293240017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,3072,0.1483288870917426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,3072,0.08825688891940647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,4096,0.141302227973938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,2560,0.13687645064459905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,2560,0.08193422026104398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,2048,0.11553866333431667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,16384,0.4201066758897569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,3584,0.1284711096021864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,2048,0.0672524438963996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,1536,0.07889333036210802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,3072,0.11380088329315186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,1536,0.05415911144680447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,1024,0.0418124430709415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,2560,0.09734755754470825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,1024,0.05475111140145195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,1536,0.07140088743633695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,2048,0.08364266819424099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,768,0.04539733462863498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,768,0.03592355714903937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,1024,0.05790577994452583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,512,0.034550223085615374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,512,0.032320890161726214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,128,0.011041777829329172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,768,0.052108443445629544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,256,0.019115555617544387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,256,0.023544889357354905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,64,0.023456000619464453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,512,0.044970668024486966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,256,0.03721155391799079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,128,0.020637333393096924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,64,0.009135999613338048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,65536,128,0.032235556178622775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,65536,32,0.007220444579919179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,65536,32,0.023444443941116333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,16384,0.17860266897413465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,65536,0.7199741999308268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,65536,0.4188302093082004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,16384,0.11987377537621392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,12288,0.13663821750217015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,12288,0.09250489208433364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,10240,0.13313244448767766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,10240,0.07976444562276204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,8192,0.09229066636827256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,8192,0.06596444712744819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,7168,0.08144978019926283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,10240,0.09548533625072902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,12288,0.11099021964603001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,7168,0.06389333142174615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,16384,0.15423556168874106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,8192,0.08024444182713826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,6144,0.08252444532182482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,6144,0.05365511112742954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,5120,0.05970399909549289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,7168,0.0750195582707723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,4096,0.04019555449485779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,5120,0.04742666747834948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,4096,0.04877333177460564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,6144,0.07010666529337566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,5120,0.06358933448791504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,3584,0.053192890352673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,3072,0.03479911221398248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,3584,0.043156445026397705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,3584,0.0370399985048506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,4096,0.0565262238184611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,3072,0.038168887297312416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,2560,0.03246488836076524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,2560,0.03172889020707872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,3072,0.05117688907517327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,2048,0.026737777723206416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,2560,0.04795022143257988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,1024,0.014891554911931356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,1536,0.04011377692222595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,2048,0.027866668171352808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,1536,0.02127555509408315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,2048,0.04383822282155355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,1024,0.03549066517088149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,1536,0.02402755618095398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,1024,0.020224000016848247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,768,0.011687111523416309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,65536,0.4669857554965549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,768,0.019259555472267997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,512,0.008903111020723978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,768,0.03363644414477878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,512,0.017882666654056974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,256,0.006244444598754247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,512,0.03129244512981839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,256,0.014477334088749357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,128,0.004026666697528628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,256,0.026670222481091816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,32,0.014743111199802823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,128,0.014009777042600842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,64,0.003751111113362842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,16384,128,0.024345777100986902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,16384,64,0.014334221680959066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,16384,32,0.004028444488843282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,16384,0.14804000324673122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,65536,0.5385973188612196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,65536,0.3685697714487712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,16384,0.10338311062918769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,12288,0.1114399962955051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,12288,0.07991377512613933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,10240,0.09311822387907241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,10240,0.06904000043869019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,8192,0.07604533433914185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,8192,0.05799111392762926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,7168,0.0665280024210612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,12288,0.10191910796695286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,16384,0.17340266704559326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,10240,0.08291466368569268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,7168,0.052143110169304736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,8192,0.07125155793295966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,6144,0.058288892110188804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,6144,0.04968355430497063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,7168,0.06657244761784871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,5120,0.04855822192298042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,5120,0.04163466559516059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,6144,0.06249155600865682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,4096,0.0403146677547031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,5120,0.05615466833114624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,3584,0.04862222075462341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,4096,0.03595733311441209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,65536,0.4143493440416124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,3584,0.035381333695517644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,4096,0.051149331861072116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,3584,0.03589066531923082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,3072,0.031066666046778362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,3072,0.03141955534617106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,2560,0.026544888814290363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,3072,0.04701155424118042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,2560,0.030198223061031763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,2048,0.02218399941921234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,2560,0.04416088925467598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,2048,0.024736889534526404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,1536,0.017432888348897297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,1024,0.03420177764362759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,2048,0.04674488968319363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,1536,0.021736888421906367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,768,0.03260088960329691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,1024,0.01257599973016315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,1536,0.037500444385740496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,1024,0.019849777221679688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,768,0.01034755590889189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,768,0.018796443939208984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,512,0.007996444072988298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,512,0.016806221670574613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,256,0.005006222261322869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,512,0.030864887767367895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,256,0.01443199978934394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,128,0.0037902221083641052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,256,0.025507556067572698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,128,0.013713777893119387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,64,0.0034515555534097883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,12288,128,0.0240320000383589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,64,0.014140443669425117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,12288,32,0.0037431112594074676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,12288,32,0.01444533301724328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,65536,0.3307822280459934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,65536,0.46743912167019314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,12288,0.0739555557568868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,16384,0.12990221712324354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,16384,0.09652533133824666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,12288,0.08800088697009617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,10240,0.08544533120261298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,16384,0.10646488931443955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,8192,0.05389777819315592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,10240,0.06380799743864271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,7168,0.05257599883609348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,7168,0.04902311166127523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,12288,0.09366222222646077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,8192,0.05997688902748955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,10240,0.07831911245981853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,6144,0.04675466815630595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,8192,0.0656133360332913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,6144,0.04471288786994087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,5120,0.03940800163480971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,5120,0.052905778090159096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,4096,0.03423110975159539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,7168,0.06232355700598823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,5120,0.03951644566324022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,6144,0.05860177675882975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,4096,0.03239999877081977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,65536,0.36755021413167316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,3584,0.0288355549176534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,3584,0.031762666172451444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,3072,0.025655999779701233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,4096,0.05940177705552843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,3072,0.029848890172110662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,3584,0.045752889580196805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,2560,0.021847999758190576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,3072,0.04490844408671061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,2560,0.0262435558769438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,2048,0.01824711097611321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,2560,0.041591111156675555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,2048,0.024313777685165405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,2048,0.03729333480199178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,1536,0.014297778407732645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,1536,0.02221244408024682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,1024,0.010466666685210334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,1536,0.03647733396954007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,1024,0.019070222973823547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,768,0.0088764445649253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,1024,0.03351733419630263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,768,0.01812355551454756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,512,0.007121777368916406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,768,0.03243111239539252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,512,0.01570222278436025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,256,0.004369777937730153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,512,0.029696888393825952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,256,0.014495111174053617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,256,0.025262221693992615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,128,0.003702222059170405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,128,0.013636444177892474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,64,0.003448888866437806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,10240,128,0.023967999551031325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,64,0.014051554931534661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,10240,32,0.003488000068399641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,10240,32,0.014180445008807711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,65536,0.37175467279222274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,65536,0.24151288138495552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,16384,0.09865778022342259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,10240,0.0658408866988288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,16384,0.07361955775154962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,12288,0.06961777475145128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,12288,0.05805955330530802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,10240,0.05084177851676941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,16384,0.09182222021950616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,12288,0.0804319977760315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,8192,0.047616001632478505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,8192,0.04319644305441114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,10240,0.06731733348634508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,7168,0.04230133361286587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,8192,0.05826222234302097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,7168,0.039650665389166936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,6144,0.03758222195837233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,5120,0.03261422117551168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,6144,0.03625599874390496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,7168,0.055445333321889244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,5120,0.033878223763571844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,6144,0.05263377891646492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,4096,0.02612000041537815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,4096,0.02863111098607381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,3072,0.020625778370433383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,3584,0.04106400079197354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,5120,0.04785689049296909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,65536,0.4614826838175456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,3584,0.02330577704641554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,4096,0.044349332650502525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,3584,0.026968000663651362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,3072,0.025407110651334126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,2560,0.01775733298725552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,2048,0.03571733170085483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,3072,0.040355556541019015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,2560,0.0229120006163915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,2048,0.014427555931939019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,2560,0.04288533329963684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,2048,0.0206675562593672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,1536,0.01149777736928728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,1536,0.019345778557989333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,1024,0.00867022242810991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,1536,0.0347724457581838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,1024,0.01797155539194743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,768,0.00741155528359943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,1024,0.03230488962597317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,768,0.01643288963370853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,512,0.006631111105283101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,768,0.031207111146714952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,512,0.014537778165605334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,256,0.004024888906213972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,128,0.023447111248970032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,512,0.026776888304286536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,256,0.013732444081041547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,8192,256,0.024970667229758367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,128,0.0034231110993358824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,128,0.01295644458797243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,64,0.0033191111352708605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,64,0.013207111093733044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,8192,32,0.0034959999223550162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,8192,32,0.013626666532622443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,65536,0.2236844433678521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,65536,0.3360079924265544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,16384,0.08790489037831624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,16384,0.0683422221077813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,12288,0.06185599830415514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,12288,0.05433600147565206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,10240,0.052367998494042285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,16384,0.08876799874835545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,10240,0.04725422130690681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,12288,0.07196888658735487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,8192,0.04236622320281135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,10240,0.06400355365541247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,6144,0.03341599967744615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,8192,0.04052444299062093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,7168,0.03755555550257365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,65536,0.28698934449089897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,7168,0.03740444448259141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,8192,0.06901244322458903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,7168,0.05310844381650289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,6144,0.03417777683999803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,6144,0.05029333300060696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,5120,0.028312000963422988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,5120,0.030975110001034204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,4096,0.023494222097926672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,5120,0.04594577683342827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,4096,0.027087110612127516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,3584,0.02109777761830224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,3584,0.025310221645567153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,4096,0.04268444577852885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,3072,0.0388080014122857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,3072,0.01847999956872728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,2560,0.02169688873820835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,2048,0.013025778035322825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,3584,0.04006844427850511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,1536,0.010536889235178629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,3072,0.023345778385798138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,2560,0.01592088904645708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,2560,0.036791112687852644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,2048,0.01999288962946998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,2048,0.03533333208825853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,1536,0.01885422236389584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,1024,0.00815555536084705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,1536,0.034513778156704374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,1024,0.017498667041460674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,768,0.006974221931563483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,1024,0.03258311086230808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,768,0.015271999769740634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,512,0.005581333405441708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,768,0.031234665049446955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,512,0.014325333966149224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,256,0.0037395556767781577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,512,0.027170666389995154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,256,0.013552889227867126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,128,0.0033733333564466904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,256,0.024971556332376268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,128,0.012822222378518848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,64,0.0033324443631701996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,7168,128,0.023596444063716467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,64,0.013134222063753339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,7168,32,0.0034248888906505373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,7168,32,0.013552000125249227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,65536,0.2789413399166531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,65536,0.2077475521299574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,16384,0.07954044474495782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,10240,0.0477439992957645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,16384,0.06232088804244995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,12288,0.05391022231843737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,12288,0.050624890459908374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,16384,0.08817244238323635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,10240,0.04438222116894192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,12288,0.07591644260618421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,8192,0.037040889263153076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,8192,0.03834844297832913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,10240,0.06541244188944499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,65536,0.2732702096303304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,7168,0.03298844562636481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,7168,0.035651554663976036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,8192,0.05687466594907972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,6144,0.02952444553375244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,7168,0.05418489045566983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,6144,0.032448000378078885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,6144,0.050922665331098765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,5120,0.028258668051825628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,4096,0.04171644316779243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,5120,0.029025776518715754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,4096,0.020834666159417894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,5120,0.04586666822433472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,4096,0.025565332836574976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,3584,0.0184888889392217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,3584,0.024209777514139812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,3072,0.016504888733228047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,3584,0.03994133406215244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,3072,0.022796443767017786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,2560,0.013987556099891663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,3072,0.03853155506981744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,2560,0.020980444219377305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,2048,0.012006222373909421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,1536,0.03412533468670315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,2560,0.036474668317370944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,2048,0.020035554965337116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,1536,0.009722666607962714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,2048,0.0346684455871582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,1536,0.01828533411026001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,1024,0.007801777786678738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,1024,0.016522667474216886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,768,0.006624889042642381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,256,0.0036951113078329298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,1024,0.03143466512362162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,768,0.01516533394654592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,512,0.005051555732885997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,768,0.02847466535038418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,512,0.014323555760913424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,512,0.02612888813018799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,256,0.013537777794731988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,128,0.0032906666811969546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,256,0.024752888414594863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,128,0.01295644458797243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,64,0.003119999956753519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,6144,128,0.023333332604832117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,64,0.01295644458797243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,16384,0.0667404466205173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,6144,32,0.003220444545149803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,6144,32,0.013369777964221107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,65536,0.23358488082885742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,12288,0.05294933252864414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,65536,0.19293422169155547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,16384,0.059486223591698535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,12288,0.04696444339222378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,10240,0.038817776574028864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,16384,0.08376444710625543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,10240,0.041781332757737905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,8192,0.03145688772201538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,12288,0.06647644440333049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,8192,0.036229332288106285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,10240,0.05934400028652615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,7168,0.02832177612516615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,7168,0.049898667467965015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,8192,0.051995555559794106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,65536,0.2667173279656304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,7168,0.033187554942237005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,6144,0.025115556187099878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,6144,0.03038311004638672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,5120,0.02164977788925171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,6144,0.04695111182000902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,5120,0.027506666051016912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,4096,0.017817777064111497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,5120,0.0436737769179874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,4096,0.023949333363109167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,3584,0.015875554747051664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,3584,0.022839999861187402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,4096,0.04038666685422262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,3072,0.013917333549923368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,2048,0.010256888965765635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,3584,0.037110222710503474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,3072,0.021723555194007024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,2560,0.011877333124478659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,3072,0.03706489006678263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,2560,0.020326221982638042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,2560,0.0354053344991472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,2048,0.019680889116393197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,1536,0.00850933293501536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,2048,0.03363733490308126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,1536,0.01815022197034624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,1024,0.006737777756320105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,1536,0.0332293344868554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,1024,0.015323556131786771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,512,0.01443555619981554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,768,0.005865777946180767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,1024,0.031209776798884075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,768,0.014846222268210517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,512,0.004242666479614046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,768,0.028565333949195013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,256,0.003448888866437806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,256,0.013600889179441663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,512,0.02644177774588267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,128,0.003228444399105178
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,256,0.024812445044517517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,128,0.012870222330093384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,64,0.003023111157947116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,5120,128,0.023367111881573994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,64,0.012815111213260226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,16384,0.05212444398138258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,5120,32,0.00315022220214208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,5120,32,0.01313066648112403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,65536,0.1882666614320543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,65536,0.18418755796220568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,16384,0.055813332398732506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,12288,0.04212888744142321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,10240,0.03963377740648057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,12288,0.04503200120396084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,16384,0.07791822486453585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,10240,0.036128888527552284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,12288,0.06435733371310763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,8192,0.030379556947284277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,8192,0.03454666667514377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,10240,0.05832088655895657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,65536,0.24162401093377006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,7168,0.027698665857315063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,8192,0.05005244414011637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,7168,0.033350222640567355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,6144,0.024889777104059856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,7168,0.0476399991247389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,6144,0.029168887270821467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,5120,0.01833866702185737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,6144,0.046352889802720815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,5120,0.02553333342075348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,5120,0.041470223002963595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,4096,0.014914666612943014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,4096,0.023247111174795363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,3584,0.013089777694808112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,4096,0.03870044483078851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,3584,0.022405332989162866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,3072,0.0118568887313207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,3584,0.036714666419559054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,2560,0.03463110989994473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,3072,0.021448888712459143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,2560,0.010447999669445885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,3072,0.03654399845335219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,2560,0.019909333851602342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,2048,0.009025777379671732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,1024,0.006154666758245892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,1536,0.03222222129503886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,2048,0.01904622217019399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,1536,0.007623111208279927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,2048,0.033048888047536217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,1536,0.017450667089886136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,512,0.0037475555307335327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,1024,0.015290665957662793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,768,0.005040888984998067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,1024,0.028274665276209515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,768,0.014692443940374585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,768,0.02721422248416477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,512,0.014392000105645923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,256,0.003294222263826264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,512,0.026174222429593403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,128,0.023211555348502264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,256,0.013340444200568728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,128,0.003043555551105075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,4096,256,0.024528000089857314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,128,0.01276622215906779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,64,0.0029120000286234748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,64,0.012859555582205454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,4096,32,0.003265777809752358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,4096,32,0.013034666577974955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,65536,0.17392888334062365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,65536,0.17190399434831408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,16384,0.04891555507977804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,16384,0.05387111173735725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,12288,0.03801777627733018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,16384,0.07587822278340657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,12288,0.04302311274740431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,10240,0.058747556474473744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,12288,0.06195377641254001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,10240,0.03263555632697211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,10240,0.03827111257447137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,65536,0.2290275626712375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,8192,0.027663111686706543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,8192,0.032772445016437106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,7168,0.021447999609841242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,8192,0.04886933167775472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,7168,0.03180888957447476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,7168,0.0460382236374749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,6144,0.01899111105336083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,6144,0.027976890405019123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,6144,0.043904887305365674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,5120,0.016231111354298063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,5120,0.024832000335057575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,4096,0.01316088851955202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,5120,0.040355556541019015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,3584,0.03567644291453891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,4096,0.02251822253068288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,4096,0.03784622086419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,3584,0.011954666839705573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,3584,0.021706667211320665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,3072,0.010833777487277985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,3072,0.02077511118517982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,3072,0.035770667923821345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,2560,0.009588444398509132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,2560,0.019448000523779128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,2048,0.008431110945012834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,2560,0.034325334760877825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,2048,0.01865244408448537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,1536,0.007272889216740926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,2048,0.0328400002585517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,1536,0.016172443826993305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,1024,0.005635555419656966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,1536,0.031369778845045306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,1024,0.015281778242852954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,768,0.004433777597215441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,1024,0.027795556518766616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,768,0.014493332968817817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,512,0.0037226668662495087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,768,0.02693422304259406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,512,0.014087999860445658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,256,0.0032906666811969546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,512,0.025655999779701233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,256,0.013437333206335703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,256,0.024453333682484094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,128,0.003026666740576426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,128,0.012812444733248817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,64,0.0028728888266616394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3584,128,0.02323377794689602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,64,0.012792000340090858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3584,32,0.002977777686383989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3584,32,0.013106666505336761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,65536,0.1436035500632392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,65536,0.16765244801839194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,16384,0.04425333274735344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,16384,0.052060445149739586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,65536,0.2249546580844455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,12288,0.03305955396758185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,12288,0.04180711176660326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,16384,0.07464710871378581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,10240,0.028524445162879095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,12288,0.06130311224195692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,10240,0.03684355484114753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,8192,0.024484443995687697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,10240,0.055283556381861367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,8192,0.031812445984946355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,7168,0.02204977803760105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,8192,0.04825599988301595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,7168,0.029530667596393164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,7168,0.04556177722083198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,6144,0.01974933346112569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,6144,0.026594666971100703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,5120,0.017111111018392775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,6144,0.04268533322546217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,5120,0.024663110574086506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,4096,0.014927110738224454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,5120,0.04010666741265191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,4096,0.02218399941921234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,3584,0.010967999696731567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,4096,0.03745511174201965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,3584,0.02146666745344798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,3072,0.00996088898844189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,3584,0.03567377726236979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,3072,0.020389333367347717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,2560,0.03417866759830051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,2560,0.008936888641781276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,3072,0.03545599844720628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,2560,0.019297776950730216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,1536,0.030433777305814955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,2048,0.007857778006129796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,2048,0.01850044396188524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,1536,0.006871999965773688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,2048,0.0321359998650021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,1536,0.015955555770132277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,1024,0.005616000129116907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,1024,0.015076445208655464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,768,0.00424177779091729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,1024,0.02823288904296027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,768,0.014725334114498563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,512,0.0037884443170494507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,768,0.027232888672086928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,512,0.01402933398882548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,256,0.0033048888047536216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,512,0.02607111136118571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,256,0.013373333546850415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,128,0.003051555612021022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,256,0.024226667152510747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,128,0.01276177747382058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,3072,128,0.02312177750799391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,64,0.002875555513633622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,64,0.012724444270133972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,3072,32,0.002944888960984018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,3072,32,0.01310400002532535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,65536,0.12329244613647461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,65536,0.15603644318050808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,12288,0.02846222122510274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,16384,0.03571822245915731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,12288,0.03955199983384874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,65536,0.2104560004340278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,16384,0.05054488778114319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,16384,0.0724817779329088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,10240,0.02497244377930959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,10240,0.03495999839570787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,12288,0.060531556606292725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,8192,0.02100533412562476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,10240,0.05700711409250895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,7168,0.019121777680185106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,8192,0.030006223254733618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,7168,0.028215110301971436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,8192,0.047376887665854565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,6144,0.01700799994998508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,7168,0.04538755615552267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,6144,0.025395555628670588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,6144,0.04233244392606947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,5120,0.012825777961148156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,5120,0.02328444520632426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,3584,0.020682666036817763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,3072,0.0092284447617001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,4096,0.010842666857772403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,5120,0.039776888158586286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,4096,0.021742223037613764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,3584,0.010090666512648264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,4096,0.03732088870472378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,2048,0.007350222104125553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,3584,0.035291555855009295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,3072,0.01995022263791826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,2560,0.008229333493444655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,3072,0.03545866741074456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,2560,0.01868799991077847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,2560,0.03418488966094123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,2048,0.016418667303191293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,2048,0.030200001266267564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,1024,0.027948445743984644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,1536,0.0063946665161185795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,1536,0.015847999188635085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,1024,0.005165333135260476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,1024,0.015012444721327888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,1536,0.029872000217437744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,768,0.0041955556306574075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,512,0.003681777666012446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,768,0.014442667365074158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,768,0.02686400049262577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,128,0.012738666600651212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,512,0.013839999834696451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,128,0.02327377763059404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,256,0.0033253334048721525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,512,0.025694222913848028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,256,0.013256889250543384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2560,256,0.024530667397711012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,128,0.003015999992688497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,64,0.002875555513633622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,64,0.012670222255918713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2560,32,0.0029226665695508323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2560,32,0.012967999610635968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,65536,0.09985599915186565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,65536,0.1524453295601739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,16384,0.02961155441072252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,16384,0.04850133260091146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,12288,0.023927999867333308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,16384,0.0688906643125746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,12288,0.038041777080959745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,65536,0.1992915603849623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,12288,0.05741955836613973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,10240,0.021054221524132624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,10240,0.03361777795685662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,8192,0.017966222431924608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,10240,0.0517502228418986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,8192,0.02894222074084812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,7168,0.01620444489849938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,8192,0.04471910993258158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,7168,0.027394667267799377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,7168,0.042711999681260854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,6144,0.014882667197121514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,5120,0.038013332419925265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,6144,0.025199999411900837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,5120,0.014387556248241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,6144,0.040248887406455144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,3584,0.020607999629444547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,5120,0.02328799996111128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,4096,0.011963555382357704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,4096,0.02123111155298021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,3584,0.010565333068370819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,4096,0.03639644384384155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,3072,0.009594666461149851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,2560,0.01808355583084954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,2048,0.006879999819729064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,3584,0.03472266594568888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,2048,0.030181331766976252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,3072,0.019337778290112812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,2560,0.007655999726719326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,3072,0.034528887934154935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,2048,0.01662400033738878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,2560,0.033324443631702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,1536,0.006038222048017714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,1536,0.015597333510716757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,1024,0.0042444442709287005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,1536,0.02964799933963352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,1024,0.0147733340660731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,768,0.0038720000949170855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,1024,0.027600889404614765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,768,0.014366222752465142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,512,0.0035582222044467926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,768,0.02678488857216305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,256,0.0031662223239739737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,512,0.013895110951529609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,256,0.013039111263222165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,128,0.0029395555870400537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,512,0.025440888272391424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,256,0.024223110742039148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,128,0.012732444538010491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,64,0.002761777697337998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,2048,128,0.02292622294690874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,64,0.012694222231705984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,2048,32,0.0028213332924577924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,65536,0.07773244380950928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,2048,32,0.013033777475357056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,16384,0.02437688906987508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,65536,0.14107378323872885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,16384,0.044475555419921875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,65536,0.18019821908738878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,16384,0.06615289052327473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,12288,0.01925511161486308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,12288,0.03532622257868449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,10240,0.01680266691578759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,8192,0.042909334103266396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,12288,0.05450400047832065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,10240,0.03155288762516446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,8192,0.014112888111008538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,10240,0.04900000161594815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,6144,0.023903111616770428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,8192,0.027671110298898485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,7168,0.013171555267439948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,7168,0.026224000586403742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,6144,0.01202755586968528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,7168,0.04076711005634732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,5120,0.010970667004585266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,5120,0.022337777747048274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,6144,0.03935022155443827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,5120,0.03821511069933573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,4096,0.01001422190003925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,4096,0.02049777739577823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,3584,0.00886222223440806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,4096,0.035571555296579994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,3584,0.01920000049802992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,3072,0.008303999900817871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,3584,0.03321866525544061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,3072,0.018187556001875136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,2560,0.0074826669361856245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,3072,0.033815999825795494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,2560,0.016771554946899414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,2560,0.030536890029907227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,2048,0.006540444576077991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,2048,0.016139555308553908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,1536,0.004892444445027245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,1024,0.014631110760900708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,2048,0.03012088934580485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,1536,0.015463110473420886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,1536,0.029102222787009344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,1024,0.004206222378545337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,768,0.0038480001191298165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,768,0.026701332794295415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,1024,0.027426666683620874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,768,0.014159111513031853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,512,0.003516444315512975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,512,0.013615111509958902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,256,0.0031600000543726813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,512,0.025234666135576036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,256,0.013152889079517789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,128,0.0029351111087534162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,256,0.024271999796231587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,128,0.012656889028019376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,65536,0.13634667131635878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,64,0.0027644443843099806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1536,128,0.022970666488011677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,64,0.012539555629094442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1536,32,0.0028799999919202593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1536,32,0.013036444783210754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,65536,0.05865244732962715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,16384,0.018004443910386827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,16384,0.04318933354483711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,65536,0.1761617792977227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,10240,0.030249777767393324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,12288,0.014198222094111972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,16384,0.06375288963317871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,12288,0.034062223301993474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,10240,0.012749333348539142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,12288,0.053751998477511935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,8192,0.011262222296661802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,10240,0.048660443888770215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,8192,0.026782222919993933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,7168,0.010596444209416708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,6144,0.03898933198716905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,8192,0.04248799880345663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,7168,0.025626666016048853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,6144,0.011920889218648275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,7168,0.040765331851111516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,6144,0.02260977857642704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,5120,0.010575110713640848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,5120,0.021179555190934077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,3584,0.01830311119556427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,4096,0.009080000221729279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,5120,0.036271111832724676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,4096,0.019295111298561096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,3584,0.008475555313958062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,4096,0.0350560016102261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,3072,0.007812444534566667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,3584,0.03120000163714091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,3072,0.017512889371977914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,2560,0.007100444700982835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,2048,0.02921066681543986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,3072,0.03138666682773166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,2560,0.016720000240537856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,2560,0.030424889591005113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,2048,0.006276444428496891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,2048,0.016015999846988253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,1536,0.005094222310516569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,1536,0.015254222684436373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,1536,0.02899466620551215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,1024,0.0041751112374994485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,1024,0.014704000618722705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,768,0.003903111235962974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,256,0.0031235555393828284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,1024,0.027580445011456806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,768,0.01423111061255137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,256,0.024093333217832778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,768,0.02661955522166358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,512,0.0035004446076022256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,64,0.002736888825893402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,512,0.0136817776494556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,512,0.025399110383457605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,256,0.01314133322901196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,128,0.0028568889117903183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,65536,0.047594666481018066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,128,0.012662221988042196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,1024,128,0.022952889402707417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,64,0.012472888661755456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,1024,32,0.0028106667515304354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,1024,32,0.012816000315878125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,65536,0.13537955284118652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,16384,0.014466666513019137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,16384,0.04187466700871786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,12288,0.012300444145997366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,10240,0.030151999659008447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,65536,0.18085600270165336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,16384,0.06005066633224487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,12288,0.03384533193376329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,10240,0.013548444542619916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,12288,0.051875554853015475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,8192,0.011881777809725868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,7168,0.024837333295080397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,8192,0.026471111509535048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,10240,0.04705244302749634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,8192,0.04156622290611267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,7168,0.011193777951929303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,6144,0.010432889064153036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,7168,0.039326220750808716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,6144,0.02290311124589708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,4096,0.018912000788582694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,5120,0.00963466697269016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,3584,0.007996444072988298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,6144,0.03754222061898973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,3584,0.018200889229774475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,5120,0.020807110601001315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,4096,0.008560889297061497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,3072,0.017500445246696472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,5120,0.033923556407292686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,4096,0.03293155630429586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,3072,0.007258666886223688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,3584,0.030629333522584703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,2560,0.006137777947717243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,3072,0.03174399998452928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,2560,0.01684177749686771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,2048,0.005468444277842839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,2560,0.03020355436537001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,2048,0.016135111451148987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,1536,0.0047662220895290375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,1536,0.015314666761292351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,2048,0.02872266703181797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,1024,0.004072000169091755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,1536,0.02892000145382351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,1024,0.014874666929244995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,768,0.003721777763631609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,1024,0.027422222826215956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,768,0.014167999227841696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,512,0.003387555480003357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,768,0.026650667190551758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,128,0.012617777619096967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,512,0.013707555830478668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,256,0.003112888791494899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,512,0.025439111722840205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,256,0.01295111080010732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,256,0.02405333353413476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,128,0.0028640000770489373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,64,0.0027377777215507296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,65536,0.13232444392310247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,64,0.012599999705950419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,16384,0.011559111376603445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,768,128,0.02275466587808397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,768,32,0.0028088889602157804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,768,32,0.012798222402731577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,65536,0.03662933243645562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,16384,0.04151555564668443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,65536,0.14653066794077554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,12288,0.010588444769382477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,16384,0.05415644579463535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,12288,0.03298844562636481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,10240,0.010890666809346942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,12288,0.04674933354059855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,10240,0.030270222160551283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,8192,0.00977244476477305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,10240,0.04286311070124308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,8192,0.02678844498263465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,7168,0.009264888862768808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,8192,0.03866844375928243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,7168,0.024255999260478552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,6144,0.009866666462686326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,7168,0.03604000144534641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,6144,0.022284444835450914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,6144,0.036116444402270846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,5120,0.00907733374171787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,5120,0.02036977807680766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,5120,0.03156444430351257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,3072,0.007329777710967594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,4096,0.008252444366614023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,4096,0.018725333942307364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,3584,0.007400888535711501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,4096,0.031256887647840716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,3584,0.018183110488785636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,3584,0.030284444491068523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,3072,0.01737511157989502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,2560,0.00600622221827507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,3072,0.030210667186313208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,2560,0.016585777203241985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,2560,0.02921066681543986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,2048,0.005408000200986862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,2048,0.015967110792795818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,1536,0.004707555390066571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,2048,0.028068443139394123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,768,0.014113777213626437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,1536,0.015224888920783997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,1536,0.028493334849675495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,1024,0.004063999901215236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,1024,0.014528888795110913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,1024,0.02699911097685496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,768,0.0036880001425743103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,512,0.0033813334173626373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,768,0.026305778159035578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,512,0.013529777526855469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,512,0.024961777859263953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,256,0.003028444531891081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,64,0.0026764445420768526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,256,0.012883555557992725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,256,0.024129778146743774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,128,0.0028320000403457214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,128,0.012518222133318583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,512,128,0.02294844388961792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,64,0.012455999851226807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,512,32,0.0026791110220882627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,512,32,0.012709333664841123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,65536,0.021872000561820135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,65536,0.13005510965983072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,16384,0.010979555547237396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,65536,0.14686311615837946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,16384,0.0403786665863461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,12288,0.009720889230569204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,16384,0.05351644423272875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,12288,0.03261333372857835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,10240,0.008934222161769867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,12288,0.046181334389580615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,10240,0.028174221515655518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,8192,0.007629333270920648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,10240,0.04134933484925164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,8192,0.02474755545457204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,7168,0.00740533322095871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,8192,0.03684088918897841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,7168,0.023342221975326538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,6144,0.006705777926577463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,7168,0.03547999925083584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,6144,0.021431111627154883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,5120,0.0069395556218094295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,6144,0.03380889031622145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,5120,0.020064888728989493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,4096,0.006523555351628198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,5120,0.03139288889037238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,4096,0.01833599971400367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,4096,0.031096001466115315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,3584,0.006831999868154526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,3584,0.017657778329319425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,3072,0.006559999866618051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,3584,0.02951644526587592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,3072,0.016935111747847665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,2560,0.005943999936183293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,3072,0.030039111773173015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,2048,0.027804444233576458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,2560,0.016427555018001132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,1024,0.003966222206751506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,2048,0.005299555758635203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,2560,0.02924088968171014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,2048,0.015643555257055495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,1536,0.004607111215591431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,1536,0.014901333385043673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,512,0.0032888888898822996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,1024,0.014346667461925082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,1536,0.02852533260981242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,256,0.0029688889367712867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,768,0.003624888757864634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,1024,0.02662311163213518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,768,0.013794666363133324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,768,0.025997334056430395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,512,0.013407111167907715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,512,0.02501511077086131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,256,0.012739555703269111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,128,0.002790222151411904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,256,0.024159111910396155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,128,0.012465777496496836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,64,0.0026213334252436957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,256,128,0.022876444790098403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,64,0.012423111332787408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,256,32,0.0026275554878844153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,256,32,0.012688000169065265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,65536,0.014720888601409065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,65536,0.1278968916998969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,16384,0.007510222494602203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,65536,0.14699644512600368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,16384,0.03959911068280538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,12288,0.007287999822033777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,16384,0.0529333319928911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,12288,0.03123733401298523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,10240,0.007050666544172499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,12288,0.04406222369935778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,10240,0.02789688772625393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,7168,0.02312711046801673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,8192,0.006730666591061487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,10240,0.03998311029540168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,8192,0.02455555564827389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,8192,0.0359173317750295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,7168,0.006584888945023219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,6144,0.0063591111037466265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,6144,0.021506667137145996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,7168,0.03501866592301263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,6144,0.033634665939542986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,3584,0.006802666518423293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,5120,0.006761777732107375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,5120,0.019863999552196927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,4096,0.006304889089531369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,3072,0.016865778300497267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,5120,0.03142666816711426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,4096,0.018365333477656048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,4096,0.031049778064092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,3584,0.017521777086787753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,3584,0.029556443293889362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,3072,0.006495999793211619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,2560,0.0059368887709246735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,3072,0.029983109898037378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,2560,0.01626399987273746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,2048,0.005239999956554837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,2560,0.02903999884923299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,2048,0.015426667200194465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,2048,0.02760266595416599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,1536,0.004611555486917496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,1024,0.02661866611904568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,1536,0.014869333969222175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,1024,0.003937777959638172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,1536,0.02794755498568217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,1024,0.014228444960382251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,768,0.003604444364706675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,768,0.013764444324705334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,512,0.003302222324742211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,768,0.02569155560599433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,256,0.0029502221279674103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,512,0.013220444321632385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,512,0.025054223007626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,256,0.012773333324326409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,256,0.023735110958417256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,128,0.0027626665929953256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,128,0.012410666379663678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,64,0.002595555658141772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,80,128,128,0.022701332966486614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,64,0.012326222327020435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,128,32,0.002593777866827117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,128,32,0.01258399999803967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,65536,0.012726222475369772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,16384,0.006586666736337874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,65536,0.12811022334628636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,16384,0.038888888226615057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,12288,0.006493333313200209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,12288,0.030971556901931763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,10240,0.006647111227114995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,10240,0.027802666028340656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,8192,0.006554666492674086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,8192,0.024558222956127588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,7168,0.006351999938488007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,7168,0.023201778531074524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,6144,0.006262222097979651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,6144,0.021347555849287245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,5120,0.006635555376609166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,5120,0.0199608885579639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,4096,0.006275555739800136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,4096,0.018290667070282828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,3584,0.006708444406588872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,3584,0.017702221870422363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,3072,0.0063902222447925145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,3072,0.016741333736313712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,2560,0.005883555445406172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,2048,0.0052871112194326185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,2560,0.016180444094869826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,768,0.013838222457302941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,2048,0.015487111277050443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,1536,0.004589333302444882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,1536,0.014924445086055331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,1024,0.003922666526503033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,1024,0.014119111829333834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,768,0.003552888830502828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,64,0.002568888788421949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,512,0.0033111110743549135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,64,0.012292444705963135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,32,0.0025635556214385563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,512,0.013345777988433838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,256,0.0029271110478374693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,256,0.012650666965378655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,64,128,0.002744888886809349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,128,0.012406222522258759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,64,32,0.012554667062229581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,65536,0.011338666909270816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,65536,0.12757689423031277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,10240,0.027832888894610937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,8192,0.006521777974234686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,16384,0.006710222197903528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,16384,0.038069334295060896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,12288,0.00647199981742435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,12288,0.03126577867401971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,10240,0.006623111251327727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,8192,0.024482667446136475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,7168,0.0064560001095136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,7168,0.02290133304066128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,6144,0.006248888870080312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,6144,0.021410667233996924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,5120,0.006574222197135289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,5120,0.0199991116921107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,4096,0.00619911112719112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,4096,0.018239999810854595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,3584,0.006744888921578725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,3584,0.01755911111831665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,3072,0.0064560001095136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,3072,0.016886222693655226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,1536,0.014760888285107084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,2560,0.00590844452381134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,2560,0.016197333733240765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,768,0.0035742223262786865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,2048,0.0052728888889153796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,2048,0.015487111277050443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,1536,0.004619555754794015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,1024,0.003937777959638172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,1024,0.014092443717850579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,128,0.0027226667023367355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,768,0.013805333111021252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,512,0.0032382222513357797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,512,0.013176889055305056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,256,0.0029279999434947968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,256,0.012783110969596438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,64,0.002567111204067866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,128,0.012336889074908363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,64,0.012311110893885294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,80,32,32,0.0025600000388092464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,80,32,32,0.012420444852775998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,16384,0.37492889828152126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,16384,0.7424977620442709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,12288,0.2888373268975152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,12288,0.5720115767584907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,10240,0.47383822335137266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,10240,0.25594666269090444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,8192,0.19816711213853624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,8192,0.37731289863586426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,7168,0.3421786626180013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,7168,0.1762462192111545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,6144,0.1534986628426446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,6144,0.28501423199971515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,5120,0.23930488692389595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,5120,0.1459546618991428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,4096,0.19179556104871961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,4096,0.10916444328096177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,8192,0.2427591217888726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,3584,0.17049688763088652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,3584,0.10156089067459106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,12288,0.35027466879950625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,7168,0.19257067309485543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,6144,0.1681520011689928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,10240,0.29284000396728516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,3072,0.08899999989403619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,16384,0.4144097699059381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,3072,0.14731378025478786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,2560,0.14494044250912136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,2560,0.07598577605353461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,2048,0.09818755255805121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,2048,0.06453955835766263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,5120,0.16444977124532065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,4096,0.13839911090003118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,1536,0.07966488599777222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,1536,0.05344177616967095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,1024,0.052689777480231396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,1024,0.04822044571240743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,768,0.0421013335386912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,768,0.041125333971447416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,512,0.030328889687856037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,512,0.038763556215498186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,512,0.04328533344798618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,256,0.01756533318095737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,256,0.030171556605233088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,768,0.04951555530230204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,128,0.010326222413116032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,128,0.027482666903071936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,1536,0.06948355833689372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,3584,0.11450578106774224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,64,0.007844444778230455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,65536,32,0.006295111030340195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,64,0.026360000173250835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,65536,32,0.02646222213904063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,2560,0.09744622310002644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,65536,0.3975999885135227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,65536,0.7143075731065539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,16384,0.1890693373150296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,16384,0.12521244419945612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,12288,0.14108533329433864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,12288,0.08748355838987563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,3072,0.1028862198193868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,10240,0.11858844757080078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,2048,0.0794524417983161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,10240,0.07538400093714397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,8192,0.09416444434059991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,8192,0.06346488661236234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,1024,0.05668355358971489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,7168,0.08257955312728882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,7168,0.0576346649063958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,256,0.0350560016102261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,6144,0.07214755482143827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,6144,0.0516480008761088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,65536,128,0.030555556217829388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,5120,0.059713780879974365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,5120,0.045682665374543935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,16384,0.14303555753495958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,4096,0.04840888910823398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,12288,0.10643821954727173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,4096,0.040375110175874494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,3584,0.04282222191492716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,3584,0.037069333924187556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,10240,0.09115733040703668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,65536,0.45040353139241535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,3072,0.03935555617014567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,3072,0.034959110948774554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,8192,0.07899910873836942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,2560,0.03190755512979295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,2560,0.03151022063361274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,7168,0.07316977447933622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,2048,0.026128000683254663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,2048,0.027929776244693335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,1536,0.023669333921538457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,1536,0.024455999334653217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,6144,0.06817599799897936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,1024,0.01496266656451755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,1024,0.02169955604606205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,5120,0.06111377477645874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,768,0.011556444068749746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,768,0.019840889506869845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,4096,0.05591111050711738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,512,0.00869333330127928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,512,0.018219555417696636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,3584,0.05308711197641161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,256,0.005984888722499211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,256,0.014223999447292753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,3072,0.04951111144489712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,128,0.0038782221575578055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,128,0.014192889134089151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,2560,0.048601776361465454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,64,0.003621333175235324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,2048,0.04177333248986138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,64,0.015104888214005364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,16384,32,0.0037804444630940757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,16384,32,0.015415999624464246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,1536,0.03898222247759501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,65536,0.31712354554070366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,65536,0.5759982003106011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,16384,0.13333511352539062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,16384,0.09032711055543687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,12288,0.10232622093624538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,12288,0.07996888955434163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,1024,0.03475200136502584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,10240,0.08563999997244941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,10240,0.0623262193467882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,768,0.03222222129503886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,8192,0.06934577888912626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,8192,0.05271200007862515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,256,0.024874667326609295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,512,0.030806223551432293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,7168,0.061047997739579946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,7168,0.0481315553188324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,6144,0.053548445304234825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,16384,128,0.02392533255947961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,6144,0.043112887276543506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,5120,0.04492355717553032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,5120,0.038281778494517006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,16384,0.11750488811069065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,4096,0.03695644603835212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,4096,0.03487022386656867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,12288,0.08730489015579224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,3584,0.03278933299912347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,3584,0.03230666783120897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,65536,0.3914186689588759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,3072,0.028966221544477675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,3072,0.03006311257680257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,10240,0.07829599910312228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,2560,0.02472266720400916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,2560,0.027535999814669292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,8192,0.06879466772079468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,2048,0.020409777760505676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,2048,0.024461333950360615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,7168,0.06427377462387085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,6144,0.05884977844026354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,1536,0.016113777955373127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,1536,0.022115555074479844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,1024,0.011346666349305047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,5120,0.0540559987227122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,1024,0.02031822171476152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,768,0.009851555857393477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,768,0.019446222318543326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,4096,0.04899377624193827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,512,0.007261333366235097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,512,0.016843555702103507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,3584,0.04755111204253303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,256,0.004502222355869082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,256,0.013714666995737286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,3072,0.04489422175619337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,128,0.003576000117593341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,128,0.014132445057233175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,2560,0.04193422198295593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,64,0.0032844444115956626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,64,0.014711111783981323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,12288,32,0.0033733333564466904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,12288,32,0.014707555373509726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,65536,0.4588479995727539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,65536,0.29847555690341526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,2048,0.03778488768471612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,16384,0.1114177836312188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,16384,0.08249866962432861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,768,0.031035555733574763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,12288,0.08568888902664185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,1536,0.035972442891862653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,1024,0.03285244438383315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,12288,0.06553510824839275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,10240,0.0714391138818529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,10240,0.05683466460969713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,8192,0.05827733543184069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,8192,0.048060443666246205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,256,0.024125332633654278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,7168,0.051289776961008705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,7168,0.044413334793514676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,128,0.02348533272743225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,6144,0.04561600089073181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,12288,512,0.029303110308117334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,6144,0.040200889110565186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,5120,0.038155555725097656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,5120,0.03594933284653558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,16384,0.0982933309343126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,4096,0.03140888942612542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,4096,0.032758222685919866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,65536,0.34809067514207626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,3584,0.028744889630211726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,3584,0.03400355577468872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,12288,0.07931911283069186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,3072,0.024544888072543677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,10240,0.0711839993794759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,3072,0.028004444307751123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,2560,0.021086222595638696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,8192,0.06277422110239665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,2560,0.025566221939192876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,2048,0.01739911072784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,2048,0.023009777069091797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,7168,0.059340443876054555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,1536,0.013745778136783175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,1536,0.02125777800877889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,6144,0.054077333874172635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,1024,0.009910222556855943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,1024,0.019497777024904888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,5120,0.050478223297331065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,768,0.008287111090289222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,768,0.018389332625601027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,4096,0.04598933458328247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,512,0.006584888945023219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,512,0.015433778365453085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,3584,0.05360355642106798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,256,0.004691555682155821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,256,0.017104888955752056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,3072,0.04235555397139656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,128,0.003469333259595765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,128,0.013792888985739814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,2560,0.03843911157713996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,64,0.0032995556377702286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,64,0.01425244410832723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,10240,32,0.003478222423129611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,2048,0.03567022085189819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,10240,32,0.014280888769361707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,65536,0.3668897681766086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,65536,0.22332888179355195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,16384,0.09013333585527207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,16384,0.06727555725309584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,1536,0.034557332595189415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,12288,0.06945155726538764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,12288,0.054803556866115995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,768,0.030656887425316706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,1024,0.032631999916500516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,10240,0.058098667197757296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,10240,0.047444442907969155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,8192,0.047375109460618764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,8192,0.04105066590838962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,512,0.02868622210290697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,7168,0.04213866591453552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,7168,0.03773244553142124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,256,0.024136889312002394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,6144,0.036863999234305486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,6144,0.03484266665246751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,10240,128,0.023676445086797077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,5120,0.03110755483309428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,5120,0.030975110001034204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,16384,0.08719644281599258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,4096,0.025968000292778015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,4096,0.02702666653527154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,65536,0.28692711724175346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,3584,0.023084445132149592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,3584,0.024862221545643274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,12288,0.07095288568072848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,3072,0.020407110452651978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,10240,0.06356000237994723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,3072,0.02307644486427307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,2560,0.01754755609565311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,2560,0.020648888415760465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,8192,0.056759999858008496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,2048,0.01442399952146742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,2048,0.01942399972014957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,7168,0.05412888858053419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,1536,0.011341333389282227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,1536,0.01794488893614875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,6144,0.04991733365588718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,1024,0.008817777865462834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,1024,0.016943999462657504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,5120,0.04731466703944736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,768,0.0074142225914531285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,768,0.014853333433469137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,3584,0.0413315561082628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,4096,0.04156444470087687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,512,0.00591466658645206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,512,0.012891555825869242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,256,0.0038008888562520346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,256,0.012319111161761813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,3072,0.03838222225507101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,128,0.003383999897374047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,128,0.012183999849690331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,2560,0.03622755408287048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,64,0.0032248888164758682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,64,0.0124195557501581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,2048,0.03382133444150289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,8192,32,0.0033857776886887024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,8192,32,0.012689777546458773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,65536,0.33485865592956543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,65536,0.2068906625111898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,16384,0.0860346688164605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,16384,0.06272355715433757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,1536,0.033519999848471746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,12288,0.06172533167733086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,12288,0.05101955599255032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,1024,0.03154044349988302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,10240,0.05215733249982198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,10240,0.04433688852522108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,768,0.02981688910060459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,8192,0.04178399840990702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,8192,0.03857510950830247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,512,0.026384888423813715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,7168,0.037031110790040754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,256,0.023718222975730896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,7168,0.03519377774662442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,6144,0.03348088926739163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,6144,0.03204800022972955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,8192,128,0.023015111684799194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,5120,0.027912888262006972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,5120,0.028772443532943726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,16384,0.08196622133255005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,4096,0.02427022159099579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,4096,0.02557955516709222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,65536,0.26586932606167263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,3584,0.021518222159809534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,3584,0.023574221465322707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,12288,0.06798666715621948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,3072,0.01903377804491255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,3072,0.021544888615608215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,10240,0.06309155623118083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,2560,0.016565332810084026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,2560,0.020291555258962844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,8192,0.05500622259245979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,2048,0.013400000002649097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,2048,0.018735110759735107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,7168,0.05191555619239807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,1536,0.01038755559258991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,1536,0.01738577749994066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,6144,0.05060266786151462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,1024,0.008294222255547842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,1024,0.01604977746804555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,5120,0.04507911205291748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,768,0.006991110742092133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,768,0.013620444469981723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,4096,0.04085777865515815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,512,0.005659555395444234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,3584,0.03974399964014689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,512,0.012851555314328937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,256,0.0036959999965296853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,256,0.012247111234400006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,3072,0.03812888926929898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,128,0.003334222154484855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,128,0.012117333710193634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,2560,0.03557599915398492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,64,0.003126222226354811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,64,0.012500444220172035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,7168,32,0.0032097777972618737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,7168,32,0.012479110724396177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,65536,0.27575733926561147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,65536,0.18968266910976836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,2048,0.03331289026472304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,16384,0.07459644476572673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,1536,0.03322666552331712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,16384,0.0578133331404792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,12288,0.058564444382985435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,1024,0.03122666809293959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,768,0.02912799848450555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,12288,0.046851555506388344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,10240,0.04970577690336439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,10240,0.04095200035307143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,8192,0.03824799921777513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,8192,0.03570222192340427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,512,0.02558311157756382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,7168,0.03379288978046841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,256,0.023802666200531855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,7168,0.032666666640175715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,6144,0.029845333761639063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,6144,0.02981333269013299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,7168,128,0.02305511136849721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,5120,0.02458133300145467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,5120,0.02686133318477207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,65536,0.2262649006313748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,4096,0.022986667023764715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,4096,0.023955555425749883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,16384,0.0754151079389784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,3584,0.018227555685573153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,3584,0.02164266672399309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,10240,0.056629333231184215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,12288,0.06321066617965698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,3072,0.016089777151743572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,3072,0.020505777663654752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,8192,0.05142933461401197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,2560,0.013903111219406128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,2560,0.019179556104871962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,2048,0.011708444191349877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,2048,0.018210666047202218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,7168,0.04914222160975138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,1536,0.009513777991135916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,6144,0.044999112685521446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,1536,0.01702133317788442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,1024,0.007584888901975419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,1024,0.015021334091822306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,5120,0.04308266772164238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,768,0.006497777584526274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,768,0.013140444126394061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,4096,0.03818577859136794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,512,0.004580444345871608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,512,0.01277955538696713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,3584,0.0377475553088718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,256,0.003576000117593341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,3072,0.03610488772392273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,256,0.012146666646003723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,128,0.0031973332580592898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,128,0.0120666664507654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,2560,0.03447555502255758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,64,0.003027555636233754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,64,0.012225777738624148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,6144,32,0.0032222221295038858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,6144,32,0.012269333004951477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,2048,0.032829334338506065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,65536,0.23121065563625762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,65536,0.16136711173587376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,16384,0.06273066997528076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,16384,0.051698668135537036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,1536,0.032646222247017756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,12288,0.049590223365359835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,12288,0.04208711120817396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,1024,0.03104089034928216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,10240,0.03843644592497084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,10240,0.03719110952483283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,768,0.028903110159768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,8192,0.03122488988770379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,8192,0.03257866700490316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,512,0.025416889124446448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,7168,0.02754844393995073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,7168,0.02997422218322754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,256,0.02350844442844391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,6144,0.02509511179394192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,6144,0.02714311083157857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,6144,128,0.02313422163327535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,5120,0.024047111471494038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,5120,0.024293333292007446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,16384,0.07002488772074382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,4096,0.01771022213829888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,65536,0.213898671997918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,4096,0.021554667088720534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,3584,0.017975111802419026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,3584,0.020208888583713107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,12288,0.05889155467351278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,3072,0.013924444715181986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,3072,0.01906755566596985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,10240,0.05353333221541511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,2560,0.012049777640236748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,2560,0.018418666389253404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,8192,0.04850666721661886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,2048,0.010145777629481422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,2048,0.017539555827776592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,7168,0.046569777859581836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,1536,0.008550222549173567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,1536,0.01649777756796943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,6144,0.042316443390316434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,1024,0.006824000014199152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,1024,0.013456000222100152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,5120,0.040192888842688665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,768,0.005909333212508096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,768,0.013057777451144325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,4096,0.037449777126312256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,3584,0.036708444356918335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,512,0.003988444391224119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,512,0.012602667013804117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,256,0.003375111147761345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,256,0.012069332930776807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,3072,0.03493244449297587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,128,0.0031031111462248694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,128,0.011874666644467248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,2560,0.03391644358634949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,64,0.0029760001020299066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,64,0.01240533341964086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,5120,32,0.003040888864133093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,5120,32,0.012261333564917246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,65536,0.18466933568318686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,65536,0.14103378189934626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,2048,0.031743112537595965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,16384,0.06300889121161567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,16384,0.046184000041749745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,1536,0.032250665956073336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,12288,0.04077600108252631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,12288,0.03817244370778402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,768,0.026174222429593403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,1024,0.02981333269013299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,10240,0.03496888942188687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,10240,0.03366488880581326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,8192,0.02929511004024082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,8192,0.02904799911710951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,512,0.025275554921891954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,256,0.02352977792421977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,7168,0.02629599968592326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,7168,0.027162666122118633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,6144,0.02111022174358368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,6144,0.024496888120969135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,5120,128,0.022986667023764715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,5120,0.018139556050300598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,5120,0.021886222892337378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,16384,0.06474044587877062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,4096,0.014797333214018079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,65536,0.1727688842349582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,4096,0.01997066703107622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,3584,0.013209777573744455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,3584,0.019276445110638935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,12288,0.05404533280266655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,3072,0.011830222275522021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,3072,0.018276444739765592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,10240,0.0492213335302141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,8192,0.044439110491010875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,2560,0.010102222363154093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,2560,0.017652443713612027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,2048,0.009125333693292405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,2048,0.016751110553741455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,7168,0.042711999681260854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,1536,0.0074720001882976955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,6144,0.03985777828428481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,1536,0.015067555838161044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,1024,0.005993777679072485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,1024,0.013207111093733044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,5120,0.03761333227157593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,768,0.00481599983241823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,768,0.012813333008024426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,3584,0.034860445393456355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,4096,0.03521155648761325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,512,0.003685333248641756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,512,0.012478222449620565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,256,0.0032666667054096856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,256,0.011950222154458364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,3072,0.033971554703182645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,128,0.0029688889367712867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,128,0.01202933324707879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,2560,0.032467554012934365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,64,0.0028435554769304064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,2048,0.03096710973315769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,4096,32,0.002895111011134254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,64,0.012186666329701742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,4096,32,0.012051555845472546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,65536,0.16201510694291857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,65536,0.1346586677763197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,16384,0.04840622345606486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,16384,0.043508443567487926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,1536,0.03168355425198873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,12288,0.03775377737151252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,12288,0.036306666003333196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,1024,0.028163555595609877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,10240,0.032320890161726214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,768,0.025846223036448162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,10240,0.03211999932924906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,8192,0.027993778387705486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,8192,0.027906666199366253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,512,0.025103999508751765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,7168,0.024524443679385718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,7168,0.025384889708624944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,256,0.023418666587935552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,6144,0.022302221920755174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,6144,0.022813333405388728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,4096,128,0.023048889305856492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,5120,0.016081778539551627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,5120,0.020798222886191476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,16384,0.061653335889180504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,4096,0.01311555587583118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,65536,0.17320888572269014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,4096,0.019370666808552213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,3584,0.01184800018866857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,3584,0.018762666318151686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,12288,0.05204000075658163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,3072,0.010828444527255164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,3072,0.018025777406162687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,10240,0.047373334566752114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,2560,0.009558222360081142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,2560,0.017238222890430026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,8192,0.04271999994913737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,2048,0.008749333520730337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,2048,0.016189333465364244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,7168,0.041181332535213895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,1536,0.007280000381999546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,1536,0.013844444519943662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,6144,0.03751377926932441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,1024,0.0058648888435628675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,1024,0.013262222210566202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,5120,0.03695199886957804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,768,0.00444533344772127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,768,0.012875555290116204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,4096,0.034033778640959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,512,0.003654222107595868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,3584,0.03432177835040622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,512,0.012627555264366997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,256,0.0032506665835777917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,256,0.012049777640236748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,3072,0.03352533446417914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,128,0.0029786665820413162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,128,0.01203555530971951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,2560,0.032250665956073336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,64,0.0028604444944196274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,64,0.011941333611806234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,2048,0.030932443009482488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3584,32,0.0028853333658642242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3584,32,0.0118568887313207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,65536,0.14302310678693983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,65536,0.1298640039232042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,16384,0.04059377643797133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,16384,0.04169244567553202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,1536,0.03123289015558031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,12288,0.033180445432662964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,12288,0.0346684455871582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,1024,0.027268444498380024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,10240,0.027970665031009253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,10240,0.030903998348448012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,768,0.02608000073168013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,8192,0.023341332872708637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,8192,0.026320889592170715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,512,0.024880000286632117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,256,0.023393778337372675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,7168,0.021388444635603163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,7168,0.024062222904629175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,6144,0.016851555969980028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3584,128,0.022967999180157978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,6144,0.021948445174429152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,5120,0.01422933406300015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,5120,0.020450666546821594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,65536,0.15361866686079237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,16384,0.058951112959120006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,4096,0.013254221942689685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,4096,0.01884622209601932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,3584,0.010899555351999072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,3584,0.01829333272245195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,12288,0.049729777706993945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,3072,0.009955555200576782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,3072,0.017691555950376723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,10240,0.04549422197871738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,2560,0.008832889298597971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,2560,0.01702044407526652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,8192,0.04116977916823493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,2048,0.007723555796676212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,2048,0.014514666464593677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,7168,0.0391813317934672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,1536,0.006776000062624614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,1536,0.013731555806265937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,6144,0.0370053317811754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,5120,0.035692443450291954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,1024,0.005254222287072076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,1024,0.013193777865833707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,768,0.003991999973853429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,768,0.012855110896958245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,4096,0.033823112646738686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,512,0.0036968886852264404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,3584,0.03403822249836392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,512,0.012474666866991254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,256,0.0032257777121331957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,256,0.011870221959220039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,3072,0.033484445677863225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,128,0.002970666728085942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,128,0.011867555479208628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,2560,0.032347556617524885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,64,0.002871111035346985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,64,0.012124444047609964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,3072,32,0.002895999906791581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,3072,32,0.011906666888131036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,65536,0.1217599974738227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,65536,0.12653866079118517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,2048,0.030718220604790583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,16384,0.034874667723973594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,16384,0.0403786665863461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,1024,0.027071111732059058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,1536,0.030581331915325586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,12288,0.028007109959920246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,12288,0.03322133421897888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,768,0.025693333811230127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,10240,0.0244186669588089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,10240,0.02937955657641093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,8192,0.02064533366097344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,8192,0.024857777688238356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,512,0.025077333052953083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,7168,0.01697688963678148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,7168,0.024303111765119765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,256,0.023376888699001733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,6144,0.014906666345066495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,6144,0.022475555539131165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,3072,128,0.02272533377011617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,5120,0.013027555412716336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,5120,0.02030933399995168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,65536,0.14717777570088705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,4096,0.01074399964676963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,4096,0.01885066595342424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,16384,0.056904892126719155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,3584,0.009855111440022787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,3584,0.018213333355055917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,12288,0.048247112168206104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,3072,0.00907199995385276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,3072,0.017525333497259352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,10240,0.043302221430672534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,2560,0.008222222328186035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,2560,0.01556799974706438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,8192,0.03966310951444838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,2048,0.00721866637468338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,2048,0.0143857780430052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,7168,0.038564443588256836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,1536,0.006317333214812809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,1536,0.013550221920013428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,6144,0.03605599866973029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,1024,0.0047600000268883174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,1024,0.013051555388503604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,5120,0.03572622272703383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,768,0.00398488880859481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,768,0.012911111116409302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,4096,0.03365866674317254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,512,0.003621333175235324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,512,0.012683555483818054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,3584,0.0337777766916487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,256,0.0032097777972618737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,256,0.01202044470442666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,3072,0.033253331979115806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,128,0.0030177777840031516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,2560,0.031806220610936485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,128,0.012111110819710625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,64,0.002879111096262932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,2048,0.028747555282380845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2560,32,0.002875555513633622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,64,0.012120000190205045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2560,32,0.011890666352377998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,65536,0.09881066613727146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,65536,0.12085066901312934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,16384,0.028907554017172918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,16384,0.038662221696641706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,1536,0.028612444798151653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,12288,0.023606222536828782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,12288,0.03137422270245022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,1024,0.026717333330048457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,10240,0.02038666605949402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,10240,0.02790488799413045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,768,0.025897777742809717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,8192,0.01739199956258138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,8192,0.024547555380397375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,512,0.024912888805071514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,7168,0.01591822173860338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,7168,0.022964444425370958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,256,0.023425777753194172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,6144,0.015456888410780164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,6144,0.02139644490347968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2560,128,0.022760000493791368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,5120,0.012658666405412885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,5120,0.019780443774329293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,16384,0.05450222227308485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,4096,0.011100444528791638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,65536,0.14179378085666233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,4096,0.01846577723821004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,12288,0.04585599899291992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,3584,0.009725333087974125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,3584,0.01777066621515486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,3072,0.008401778009202745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,3072,0.016512889001104567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,10240,0.04203111264440748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,2560,0.00756177802880605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,2560,0.014852444330851236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,8192,0.03827111257447137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,2048,0.006807111203670502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,2048,0.014456000592973499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,7168,0.03770577907562256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,1536,0.005513777749405966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,1536,0.013746666411558786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,6144,0.03526488939921061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,1024,0.004480000171396467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,1024,0.013101333545313941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,5120,0.034739553928375244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,768,0.0037866665257347953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,768,0.012867555850081973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,4096,0.03286755416128371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,512,0.0034942221310403612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,3584,0.03349511159790887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,512,0.012628444366984896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,256,0.0030951110853089225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,256,0.01200977795653873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,3072,0.032604446013768516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,128,0.002854222224818336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,128,0.011778666741318174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,2560,0.030763556559880573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,64,0.002761777697337998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,2048,0.027616888284683228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,64,0.011944000091817645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,2048,32,0.0028044444819291434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,2048,32,0.011685333318180509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,65536,0.07558221949471368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,65536,0.11439288987053765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,16384,0.02390488816632165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,16384,0.03602044449912177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,1536,0.02795733345879449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,12288,0.018801778554916382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,12288,0.030198223061031763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,1024,0.027035555905765955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,10240,0.016539555456903245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,768,0.02576444380813175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,10240,0.02694844371742672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,8192,0.014091556270917257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,8192,0.02365333338578542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,512,0.024811555941899616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,7168,0.013209777573744455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,256,0.023399111297395494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,7168,0.02223555578125848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,2048,128,0.022854222191704646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,6144,0.011956444217099084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,6144,0.02072355482313368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,5120,0.011190222369299995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,5120,0.019315555691719055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,65536,0.1371271080440945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,4096,0.0099973330895106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,4096,0.017839999662505258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,16384,0.05195377932654487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,3584,0.00850666645500395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,3584,0.016899555921554565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,12288,0.04437333345413208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,3072,0.007908444437715743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,3072,0.015135111080275642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,10240,0.04152266515625848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,2560,0.007272889216740926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,2560,0.014860444598727755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,8192,0.03795199924045139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,2048,0.006112888869312074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,2048,0.013933334085676404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,7168,0.03720622261365255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,1536,0.005240889059172736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,5120,0.03458933366669549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,6144,0.03472444415092468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,1536,0.01350044459104538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,1024,0.004294222427739037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,1024,0.013047110703256396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,768,0.0037546666959921518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,768,0.012739555703269111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,4096,0.032596445745891996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,512,0.0034453332838084963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,512,0.012368888490729861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,3584,0.03302044338650174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,256,0.0030737777964936364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,256,0.011807999677128263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,3072,0.030252443419562444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,128,0.0028400001012616684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,128,0.011944889194435544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,2560,0.028936889436509874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,64,0.002704888789190186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,64,0.011861333416567909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1536,32,0.0027155555370781156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,2048,0.027436445156733196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1536,32,0.011619555453459421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,65536,0.052617778380711876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,65536,0.08294933372073703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,16384,0.017608000172509086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,16384,0.027464000715149775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,1536,0.027892443868849013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,12288,0.013972444666756524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,1024,0.026501332720120747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,768,0.02565866708755493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,12288,0.023948444260491267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,10240,0.01260444439119763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,10240,0.02197866638501485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,8192,0.010944888823562197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,8192,0.019759999381171334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,512,0.024792888098292883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,7168,0.010571555958853828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,7168,0.019152889649073284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,256,0.023203555080625746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,6144,0.010658666491508484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,6144,0.01805688937505086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1536,128,0.02288177775012122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,5120,0.00980088859796524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,5120,0.01718222267097897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,16384,0.05057511064741346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,4096,0.008924444516499838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,65536,0.1336488856209649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,4096,0.015055111712879606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,3584,0.008525333470768398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,3584,0.014633778068754407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,12288,0.04382755690150791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,3072,0.007824000385072496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,3072,0.014261333478821648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,10240,0.04048444496260749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,2560,0.006631111105283101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,8192,0.037096887826919556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,2560,0.01407999959256914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,2048,0.005504000104135937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,2048,0.013639999760521783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,7168,0.03694399860170152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,1536,0.004795555439260271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,1536,0.013193777865833707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,6144,0.034528887934154935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,5120,0.03451555636194017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,1024,0.004029333177540037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,1024,0.012808889150619507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,768,0.0037697777152061462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,768,0.01254488858911726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,4096,0.030960000223583643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,512,0.0033893332713180115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,3584,0.030285331938001845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,512,0.012236444486512078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,256,0.00305244450767835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,256,0.01185955521133211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,3072,0.02941688895225525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,128,0.002834666727317704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,128,0.011697777443461947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,2560,0.028479999966091577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,64,0.002716444432735443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,64,0.011540444360838996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,1024,32,0.002752888947725296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,1024,32,0.011315555208259158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,65536,0.045294221904542714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,65536,0.07926311095555623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,2048,0.0273964438173506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,16384,0.014706666270891825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,16384,0.027328888575236004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,1536,0.02767466670937008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,12288,0.012496000362767113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,12288,0.023357333408461675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,768,0.02565866708755493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,1024,0.026429333620601233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,10240,0.011510222322411008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,10240,0.02163644466135237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,8192,0.0103493332862854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,8192,0.01957244508796268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,512,0.024698666400379602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,7168,0.010057777994208867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,7168,0.01869333287080129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,256,0.023249778482649062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,6144,0.009118222528033787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,1024,128,0.02274666726589203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,6144,0.01777688827779558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,5120,0.008611555728647444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,5120,0.015676443775494892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,65536,0.13289244969685873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,4096,0.007770666645632849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,4096,0.014917333920796713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,16384,0.05028888914320204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,3584,0.008007999923494127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,3584,0.01462933255566491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,12288,0.04310133390956455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,3072,0.007120889094140794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,3072,0.01422133379512363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,10240,0.04047200083732605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,2560,0.00646666685740153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,2560,0.013864888913101621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,8192,0.03703822361098395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,2048,0.005384888913896348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,2048,0.013437333206335703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,7168,0.03648266527387831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,1536,0.0047128887640105355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,1536,0.013005333642164866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,6144,0.03293155630429586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,1024,0.004015111260943943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,5120,0.033987555238935686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,1024,0.012702222499582501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,768,0.0037253333462609183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,768,0.012528888881206512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,4096,0.030180444320042927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,512,0.0033902221669753394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,512,0.012128888732857175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,3584,0.031261334816614784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,256,0.003032888803217146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,256,0.011819555527634092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,3072,0.02964977754486932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,128,0.002842666581273079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,128,0.011649777491887411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,2560,0.028661333852344092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,64,0.002704888789190186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,64,0.01166488892502255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,768,32,0.0027555556346972785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,768,32,0.011496889094511667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,2048,0.027297778262032404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,65536,0.03362577822473314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,65536,0.076164444287618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,16384,0.011519999967681037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,16384,0.02664533257484436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,1536,0.027500443988376196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,12288,0.010223111344708337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,1024,0.02678844498263465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,12288,0.023001778456899855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,10240,0.00927022182279163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,10240,0.02127733329931895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,768,0.02537333303027683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,8192,0.008756444685988957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,8192,0.018231110440360177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,512,0.02477866742346022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,7168,0.00849333322710461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,256,0.023352000448438857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,7168,0.017305778132544625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,6144,0.007658667034573025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,6144,0.01613333324591319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,768,128,0.022781333989567224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,5120,0.0075804442167282104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,5120,0.015504888362354703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,65536,0.13258666462368435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,4096,0.0069297779765393995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,4096,0.014760888285107084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,16384,0.049865778949525624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,3584,0.007241778075695038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,3584,0.014490667316648694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,12288,0.04305866691801283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,3072,0.0068408888247278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,3072,0.013904000322024027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,10240,0.04025333457522922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,2560,0.005972444597217772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,2560,0.013527111046844058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,8192,0.0370488895310296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,2048,0.0052897776994440295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,2048,0.013271110753218332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,7168,0.034713778230879046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,1536,0.004691555682155821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,6144,0.03301422132386102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,1536,0.01297511077589459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,1024,0.004007110993067424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,1024,0.012620444099108378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,5120,0.03177333209249709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,768,0.003638222399685118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,768,0.012338666452301873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,4096,0.02995377779006958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,512,0.0033377777371141645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,3584,0.03026311265097724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,512,0.012039999994966718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,256,0.002992000016901228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,256,0.011791110866599612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,3072,0.02964622113439772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,128,0.002797333316670524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,128,0.011683555940786997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,2560,0.028211557202868994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,64,0.002625777696569761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,64,0.011538666983445486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,512,32,0.002628444383541743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,512,32,0.01126844435930252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,65536,0.0225591113169988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,65536,0.0731733308898078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,2048,0.027041777968406677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,16384,0.008824889030721452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,16384,0.02604177759753333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,1536,0.0277511113219791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,12288,0.007907555335097844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,12288,0.022239999638663396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,1024,0.026524444421132404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,10240,0.007675555845101674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,10240,0.020062221421135795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,768,0.025470222036043804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,8192,0.007137777904669444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,8192,0.017988445030318368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,512,0.024446222517225478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,7168,0.006937777830494775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,7168,0.016935111747847665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,256,0.02330222229162852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,6144,0.00666844430896971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,6144,0.016155555844306946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,512,128,0.022845332821210224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,5120,0.006829333388143116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,5120,0.015378667248619927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,65536,0.13378755251566568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,4096,0.006426666759782367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,4096,0.01463644372092353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,16384,0.04976088802019755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,3584,0.006919999917348226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,3584,0.014295111099878946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,12288,0.0425875551170773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,3072,0.006554666492674086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,10240,0.03889866669972738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,3072,0.013914666242069669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,2560,0.005984888722499211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,2560,0.013574221895800697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,8192,0.03516888949606154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,2048,0.0053137776752312975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,2048,0.013380444712109037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,7168,0.03412355648146735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,1536,0.00462311092350218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,6144,0.03207111027505662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,1536,0.012852444416946836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,1024,0.003968888686762916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,1024,0.012431999875439538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,5120,0.03165866600142585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,768,0.003607999947335985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,768,0.012320000264379712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,4096,0.029509332444932725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,512,0.0033173333439562055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,3584,0.030008888906902734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,512,0.012002666791280111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,256,0.0029564443975687027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,256,0.011783999701340994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,3072,0.02943199872970581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,2560,0.028000887897279527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,128,0.002777777819169892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,128,0.011607999602953592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,64,0.0026586666289303037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,64,0.011531555818186866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,256,32,0.0026106666773557663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,256,32,0.011225777367750803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,65536,0.013968888256284924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,65536,0.07388533486260308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,2048,0.026643556025293138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,16384,0.00692622239391009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,16384,0.02477333280775282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,1536,0.027444443768925134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,1024,0.026191110412279766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,12288,0.006801777829726537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,12288,0.020901333954599168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,10240,0.006731555693679386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,10240,0.019261333677503798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,768,0.02537155482504103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,8192,0.006617777877383762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,8192,0.017662222186724346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,512,0.02457688914404975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,7168,0.00647288892004225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,7168,0.016907556189431083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,256,0.023099554909600153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,6144,0.006330666856633292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,6144,0.016169778174824186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,256,128,0.022655111220147874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,5120,0.006650666809744305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,5120,0.015232000086042615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,65536,0.13392799430423313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,16384,0.048355556196636625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,4096,0.006260444306664997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,4096,0.014698666003015308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,3584,0.006718222051858902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,3584,0.014314666390419006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,12288,0.04151111178927951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,3072,0.006441777778996362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,3072,0.013972444666756524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,10240,0.03767022159364488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,2560,0.005945777727497949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,2560,0.013516444298956128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,8192,0.0344471103615231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,2048,0.00526577772365676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,2048,0.013265777793195514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,7168,0.03397688931889004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,1536,0.004602666530344221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,1536,0.012905778156386482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,6144,0.031952887773513794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,1024,0.003992889076471329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,1024,0.012636444634861417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,5120,0.03135111265712314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,768,0.003591999825504091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,768,0.012343111137549082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,4096,0.02937155630853441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,3584,0.02994222111172146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,512,0.003272888975010978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,512,0.012175111307038201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,256,0.0029057777590221832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,256,0.011685333318180509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,3072,0.029514667060640123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,128,0.0027582223216692605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,128,0.011588444312413534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,2560,0.028024001253975764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,64,0.0026168889469570587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,64,0.011374221907721626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,128,32,0.0026231110095977783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,128,32,0.01108977778090371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,65536,0.01310488912794325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,65536,0.073853333791097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,16384,0.006527110934257507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,16384,0.0244835548930698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,12288,0.0064088888466358185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,12288,0.020861332615216572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,10240,0.006565333240562015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,10240,0.019163555569118924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,8192,0.0064897777305708984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,8192,0.017694221602545846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,7168,0.0063324446479479475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,7168,0.01685333251953125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,6144,0.006173333360089197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,6144,0.015967999895413715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,5120,0.006525333142942852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,5120,0.015248888068728976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,4096,0.006157333238257303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,4096,0.014568888478808932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,3584,0.00657066661450598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,3584,0.014217777384652032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,3072,0.006275555739800136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,3072,0.013861333330472311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,2560,0.005943999936183293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,2560,0.013533333109484779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,2048,0.0053119998839166425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,2048,0.013222222526868185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,1536,0.004595555365085602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,1536,0.012830221818553077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,1024,0.003914666672547658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,1024,0.012446222205956778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,768,0.0035786665976047516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,2048,0.02682755556371477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,768,0.012271111210187277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,512,0.003244444520937072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,256,0.002903999967707528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,512,0.012048000262843238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,128,0.0027546667390399506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,256,0.011615999870830111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,128,0.011555555793974133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,64,0.00259911103381051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,64,0.01144266708029641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,64,32,0.002556444456179937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,64,32,0.0110880004035102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,65536,0.011590222517649332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,16384,0.00646577775478363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,65536,0.07412088579601712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,16384,0.02437955637772878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,12288,0.0063528890411059065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,12288,0.020890666378868952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,10240,0.006583999842405319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,8192,0.0064080001579390625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,10240,0.01921955578856998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,8192,0.017733333839310538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,7168,0.006284444282452266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,7168,0.016851555969980028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,6144,0.006114666660626729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,1536,0.027452444036801655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,6144,0.01609422266483307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,5120,0.006462222172154321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,4096,0.006051555689838197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,5120,0.015233778291278414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,4096,0.014533332652515836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,3584,0.006401777681377199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,3072,0.006075555665625467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,3584,0.014263111684057446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,3072,0.013854222165213691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,2560,0.0059075554211934405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,2560,0.01350844403107961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,2048,0.005209777918126848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,1536,0.004523555437723796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,2048,0.013197333448463015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,1024,0.003908444609906939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,1536,0.01279022213485506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,768,0.003552000141806073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,1024,0.012411555482281579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,512,0.003237333355678452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,768,0.012254222399658628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,256,0.0028728888266616394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,512,0.011911999848153857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,128,0.002691555561290847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,256,0.011648889217111798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,128,0.011403555671374003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,1024,0.026327111654811438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,64,0.002534222271707323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,64,32,32,0.00254844439526399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,64,0.011355555719799466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,64,32,32,0.011156444748242697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,16384,0.36987733840942383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,16384,0.7184462017483182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,12288,0.5456310907999674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,12288,0.2844524383544922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,768,0.025073778298166063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,10240,0.23904710345798066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,10240,0.4562266667683919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,512,0.024322667055659827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,8192,0.19578754901885986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,8192,0.3872062100304498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,256,0.023004444109068975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,7168,0.17388621966044107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,7168,0.3183777862124973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,6144,0.27517422040303546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,6144,0.1701493395699395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,5120,0.22927644517686632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,5120,0.14396088653140598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,10240,0.29312533802456325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,12288,0.34679378403557676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,8192,0.24139822853936088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,7168,0.21472533543904623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,64,128,128,0.022750222020679053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,6144,0.18800089094373915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,4096,0.1873342196146647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,4096,0.11949866347842747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,3584,0.16301066345638698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,3584,0.09783999787436591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,3072,0.14124266306559244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,3072,0.08727733294169109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,5120,0.16121689478556314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,2560,0.1179822285970052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,16384,0.4124746587541368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,2560,0.08232800165812175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,2048,0.0953928894466824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,2048,0.06386311186684503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,4096,0.1343244446648492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,1536,0.07379200061162312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,1536,0.057067553202311196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,3584,0.12442133161756729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,3072,0.10339199834399754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,1024,0.04298133320278592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,1024,0.05080088973045349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,2560,0.09085688988367717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,768,0.04060622387462192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,512,0.028789334826999243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,768,0.03932977716128031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,2048,0.08752177821265326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,1536,0.07121866941452026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,1024,0.05518044365776909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,512,0.03515999846988254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,768,0.0485155549314287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,256,0.016339555382728577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,512,0.041737778319252856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,256,0.030527111556794908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,128,0.00979466653532452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,256,0.033759110503726535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,128,0.026871111657884385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,64,0.007339555356237624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,65536,128,0.030223998758527968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,64,0.02539644473128849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,65536,32,0.0058106668293476105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,65536,32,0.026537777649031744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,16384,0.17542844348483613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,65536,0.3857662147945828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,16384,0.10802488856845432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,12288,0.13370845052931044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,65536,0.7082853317260742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,12288,0.08584355645709568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,10240,0.11176266935136582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,10240,0.07370399766498141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,8192,0.09038311243057251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,8192,0.06207288636101616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,7168,0.07967999908659193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,7168,0.05660711394415962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,6144,0.06980533070034452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,12288,0.11221155855390762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,6144,0.05057511064741346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,10240,0.09909600019454956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,5120,0.05830044216579861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,8192,0.08223644230100843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,5120,0.0454382234149509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,4096,0.047657777865727745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,7168,0.07206755876541138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,4096,0.03949777616394891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,6144,0.07118933068381415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,3584,0.0422346658176846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,3584,0.03667644328541226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,5120,0.06102666589948866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,3072,0.033773332834243774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,16384,0.1804577774471707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,3072,0.036906666225857206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,4096,0.05468800001674228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,3072,0.04804889029926724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,3584,0.051790220869912036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,2560,0.03164711263444688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,1536,0.02054222259256575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,2560,0.03091111116939121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,2048,0.02587466604179806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,2048,0.027821332216262817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,2560,0.05909511115815905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,1024,0.021054221524132624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,2048,0.04275111026234097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,768,0.012190221912331052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,1536,0.024661334024535284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,1024,0.014518222875065274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,1536,0.03800266649987962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,768,0.019690665933820937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,1024,0.03389422098795573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,512,0.008949333594905006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,512,0.018181333939234417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,768,0.03196177879969279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,256,0.00598933340774642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,512,0.02972444560792711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,256,0.014155555102560254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,128,0.0038968887594011095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,128,0.01424888935354021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,256,0.02409066590997908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,64,0.0034888887570963967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,65536,0.4366773234473334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,16384,128,0.023349333140585158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,64,0.01479200025399526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,16384,32,0.00360000009338061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,65536,0.3075173430972629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,65536,0.5526586638556586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,16384,32,0.016063110695944894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,16384,0.15446578131781682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,16384,0.08762222528457642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,12288,0.10884266429477268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,12288,0.07673599984910753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,10240,0.09124266439014012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,10240,0.06268622477849324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,8192,0.07467910978529188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,8192,0.051423112551371254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,16384,0.1039208902253045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,7168,0.06892533434761895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,12288,0.0950088898340861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,6144,0.05674488676918877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,10240,0.07718311415778266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,7168,0.04682755470275879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,6144,0.04353333181805081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,5120,0.04738755689726936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,8192,0.09895110792583889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,5120,0.038584887981414795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,4096,0.033944000800450645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,7168,0.06275200181537204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,4096,0.03919822308752272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,6144,0.0580364465713501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,3584,0.03455200129085117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,5120,0.05445333321889242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,3584,0.03183822168244256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,3072,0.029640889830059473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,3584,0.046872887346479625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,4096,0.06459199719958834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,3072,0.030167112747828167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,2560,0.0258240004380544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,2560,0.027165333429972332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,3072,0.04382755690150791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,2048,0.022250667214393616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,2560,0.0422053337097168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,1536,0.021984888447655573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,2048,0.02410488824049632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,1536,0.016827555166350473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,2048,0.038314667012956403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,65536,0.37909332911173504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,1024,0.012129777835475074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,1536,0.03738311264250014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,512,0.007673777639865875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,768,0.031594667169782854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,1024,0.019817777805858187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,768,0.009836444424258338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,1024,0.03223111232121786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,256,0.01405422223938836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,512,0.028744889630211726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,768,0.018868444694413077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,512,0.016819554898473952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,256,0.005218666460778978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,128,0.003669333126809862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,256,0.023906666371557448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,128,0.013737777868906656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,64,0.0032248888164758682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,64,0.014273777604103088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,12288,128,0.023518222901556227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,12288,32,0.0033315554675128725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,65536,0.2734808921813965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,12288,32,0.015028445257080926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,16384,0.11341244644588894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,16384,0.0794337789217631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,65536,0.45861064063178164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,12288,0.0866853329870436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,12288,0.06856622298558553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,10240,0.08236266507042779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,10240,0.05558400021659004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,8192,0.059319999482896596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,8192,0.047255999512142606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,16384,0.09495200051201715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,7168,0.05209155546294319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,12288,0.07916711436377631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,6144,0.04567644331190321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,7168,0.045838223563300244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,10240,0.07075288560655382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,6144,0.039706667264302574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,7168,0.058520886633131236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,8192,0.06174400117662218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,5120,0.038550221257739596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,5120,0.03726844324005975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,4096,0.03184088733461168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,4096,0.03141244583659702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,5120,0.05280444357130262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,6144,0.05407555566893684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,3584,0.028185778194003638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,4096,0.045979556110170156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,3584,0.03189333279927572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,3072,0.02497244377930959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,3072,0.041399998797310725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,3072,0.02737155556678772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,3584,0.04372177852524651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,2560,0.021316443880399067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,2560,0.0247999992635515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,2048,0.017794667018784415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,2048,0.02300888962215847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,65536,0.32600710127088756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,2560,0.03952444593111674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,1536,0.013776889277829064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,2048,0.036350223753187395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,1536,0.02127555509408315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,768,0.018067555295096505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,1024,0.010218666659461128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,1536,0.03435466686884562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,512,0.015849777393870883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,1024,0.01928622192806668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,768,0.008802666432327693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,1024,0.032050665881898664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,512,0.006791111081838608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,768,0.030681778987248737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,256,0.0040702223777771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,256,0.013714666995737286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,512,0.026965333355797663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,64,0.014107555150985718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,128,0.0035128887328836652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,256,0.02362488872475094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,128,0.013644444445768992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,64,0.0031991110493739447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,10240,128,0.023416888382699754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,10240,32,0.0032026666320032547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,10240,32,0.014564444621404013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,65536,0.21623733308580187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,65536,0.3617706563737657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,16384,0.08976444270875718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,16384,0.06474666463004218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,12288,0.06844177511003283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,12288,0.05494222376081678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,10240,0.05786933501561483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,10240,0.05092888739373949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,16384,0.08198133442136976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,8192,0.04778133498297798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,8192,0.03968177901373969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,12288,0.07515733109580146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,7168,0.04182666540145874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,7168,0.036899556716283165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,10240,0.06344266732533772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,6144,0.03708800011210971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,8192,0.05583733320236206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,6144,0.03314844436115689
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,5120,0.031239112218221027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,7168,0.05309955610169304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,5120,0.030082666211658057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,4096,0.02640977833006117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,6144,0.06074755721622043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,4096,0.025755556093321905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,5120,0.046727110942204796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,3584,0.02295822236273024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,4096,0.04218133290608724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,3584,0.024061333802011278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,3072,0.02030044462945726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,65536,0.2743164433373345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,3072,0.0226275556617313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,3584,0.04922399918238322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,2048,0.014321777555677625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,2048,0.019288889235920377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,2560,0.017433777451515198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,2560,0.02057422200838725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,3072,0.03792622354295518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,2560,0.03707911239729987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,1536,0.01202755586968528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,1536,0.017812444104088675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,2048,0.03478399912516276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,1024,0.008611555728647444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,1536,0.03333066569434272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,1024,0.016328000360065036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,512,0.012710221939616732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,768,0.007259555160999298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,1024,0.030561778280470107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,768,0.014358222484588623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,512,0.0058088890380329555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,768,0.02944799926545885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,128,0.012190221912331052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,256,0.003687111039956411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,64,0.003173333282272021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,512,0.025437333517604407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,256,0.012266666524940066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,128,0.0033733333564466904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,256,0.02346311178472307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,8192,128,0.022986667023764715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,64,0.01258666647805108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,65536,0.20043555895487467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,8192,32,0.0032346666687064697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,8192,32,0.013100444442696042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,16384,0.08539022339714898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,65536,0.3427244557274713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,16384,0.06031911240683662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,12288,0.06146666738722059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,10240,0.04304888844490051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,12288,0.04926044411129422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,8192,0.04203200009134081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,10240,0.05160355567932129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,12288,0.06643999947441949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,16384,0.10350666443506877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,8192,0.040411555104785495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,7168,0.03731822305255466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,7168,0.03400266501638625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,10240,0.0606915553410848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,6144,0.035954667462242976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,8192,0.053454223606321544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,6144,0.03126666612095303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,7168,0.05150844322310554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,5120,0.028055111567179363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,4096,0.024990222520298425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,5120,0.02836266822285122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,6144,0.047079109483295016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,3584,0.023020444644822016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,4096,0.023267555567953322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,5120,0.04836977852715386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,3072,0.021165332860416834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,3584,0.020853334003024634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,65536,0.2582284344567193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,3072,0.036399109496010676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,4096,0.04107022285461426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,2048,0.013073777986897362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,3072,0.01848088867134518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,3584,0.03831555445988973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,2560,0.016193777322769165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,2560,0.01998133295112186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,2048,0.0184906671444575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,2560,0.03601422243648105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,1536,0.010615111225181155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,2048,0.034607112407684326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,1536,0.017304889029926725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,1024,0.008147555920812819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,1536,0.032997333341174655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,1024,0.01588800052801768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,1024,0.030093332131703694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,768,0.007560888926188151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,768,0.013544888959990608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,512,0.0052728888889153796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,768,0.026573333475324843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,512,0.012682666381200155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,256,0.003535110917356279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,512,0.024970667229758367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,256,0.012108444339699216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,128,0.0032115555885765287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,256,0.023499555057949487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,128,0.012064889073371887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,64,0.0030133333057165146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,7168,128,0.02293066680431366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,64,0.01236088905069563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,7168,32,0.0030951110853089225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,65536,0.27290309800042045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,7168,32,0.013081777426931592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,16384,0.05526311198870341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,65536,0.18029599719577363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,12288,0.05624000231424967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,16384,0.07894666989644368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,12288,0.04513600137498644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,10240,0.04883111185497708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,10240,0.04334133201175266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,8192,0.039172444078657366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,16384,0.07180088758468628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,12288,0.06225688589943779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,8192,0.034612443712022566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,7168,0.0344124436378479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,7168,0.03199733297030131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,6144,0.029208001163270738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,10240,0.05638044410281711
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,8192,0.049943112664752536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,5120,0.026569777064853247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,6144,0.030047112041049536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,7168,0.04829511046409607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,5120,0.025677333275477093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,3584,0.019350222415394254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,6144,0.04518489042917887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,4096,0.03908177879121568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,4096,0.02162577708562215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,3072,0.01699822147687276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,4096,0.023258666197458904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,65536,0.2228026654985216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,5120,0.04266666703754001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,2560,0.019055111540688407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,3072,0.03526577684614394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,3584,0.021635555558734473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,3072,0.02014666630162133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,3584,0.037343111303117536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,2560,0.014644443988800049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,2048,0.012364444633324942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,2560,0.036215109957589045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,2048,0.018812444474962022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,1024,0.014506666196717156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,1536,0.010064000056849586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,2048,0.03345955411593119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,1536,0.01714933415253957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,1024,0.007638221813572778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,1536,0.03258488906754388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,768,0.006447111152940326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,1024,0.029503110382292006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,768,0.013257777525318993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,512,0.004655999855862724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,128,0.0031893334041039148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,512,0.012684444586435953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,768,0.028689778513378564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,256,0.0034888887570963967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,512,0.024940444363488093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,256,0.012087999946541257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,256,0.02348444362481435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,32,0.012853333519564735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,128,0.01187111106183794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,64,0.00301155551440186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,6144,128,0.022944889134830896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,6144,64,0.012288000020715924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,6144,32,0.0030586665703190696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,65536,0.22684976789686415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,65536,0.15244355466630724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,16384,0.06388888756434123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,16384,0.04885066548983256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,12288,0.04506044586499532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,16384,0.06625333097245958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,12288,0.0406275557147132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,10240,0.03829955392413669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,10240,0.035967999034457736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,12288,0.05764089028040568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,8192,0.031256887647840716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,8192,0.03170311119821336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,10240,0.053322666221194796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,7168,0.02797599964671665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,6144,0.02682400080892775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,7168,0.04562311040030586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,7168,0.028935111231274072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,8192,0.04783822099367777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,5120,0.02374222212367588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,6144,0.02494311167134179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,65536,0.19768267207675508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,5120,0.021373333202468023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,6144,0.042895111772749156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,4096,0.017647999856207106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,4096,0.020964443683624268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,5120,0.040727999475267194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,3584,0.015736889508035447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,4096,0.037984000311957464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,3584,0.021184888150956895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,3072,0.013926222920417786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,3584,0.036257776949140764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,3072,0.019076445036464267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,2560,0.013635555903116861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,2048,0.017369776964187622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,2560,0.018830221560266282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,3072,0.03418933351834615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,1536,0.01660355594423082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,2048,0.010303110712104374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,2560,0.034656888908810086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,1536,0.008575110799736446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,2048,0.03310844302177429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,1024,0.006676444576846228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,768,0.01311466677321328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,1536,0.03242399957444932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,1024,0.01348977784315745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,768,0.005980444451173146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,1024,0.02723466687732273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,512,0.004234666625658671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,768,0.026869333452648584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,512,0.012551999754375882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,256,0.00342399999499321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,256,0.011976000335481433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,512,0.02499377727508545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,64,0.012168000141779581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,128,0.003093333293994268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,256,0.02341777748531765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,128,0.011963555382357704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,64,0.0029813332690132987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,5120,32,0.0029928889125585556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,16384,0.05032533407211304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,5120,128,0.023036445180575054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,5120,32,0.01276266657643848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,65536,0.1816542281044854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,65536,0.13314310709635416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,16384,0.04368622104326884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,12288,0.04021777709325155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,12288,0.03614933292071024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,10240,0.039022223816977605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,16384,0.06070488691329956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,8192,0.028263999356163874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,10240,0.03242844343185425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,12288,0.05685422155592176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,8192,0.028726223442289565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,10240,0.049104889233907066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,7168,0.026030222574869793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,7168,0.026145777768558923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,6144,0.02312266661061181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,65536,0.1674951050016615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,8192,0.043882668018341064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,6144,0.02122933334774441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,7168,0.04255911045604282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,4096,0.01973600023322635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,5120,0.017856889300876193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,6144,0.040403554836908974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,5120,0.021287999219364587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,4096,0.014653333359294467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,5120,0.03758044375313653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,3584,0.03424711028734843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,3584,0.013223999904261695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,3584,0.019158222609096102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,4096,0.03707999984423319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,3072,0.011725333001878528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,3072,0.01815466748343574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,2560,0.010253333383136326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,2560,0.017301332619455125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,3072,0.03342933456103007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,2048,0.00886577781703737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,2048,0.01660355594423082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,2560,0.03329955538113912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,1536,0.007592889169851939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,2048,0.03243733445803324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,1536,0.015146666102939181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,1024,0.006155555446942647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,1536,0.031082666582531396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,1024,0.013329777452680798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,768,0.0047271110945277745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,256,0.003253333270549774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,1024,0.02754488918516371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,768,0.012845333251688214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,512,0.0037084445357322693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,768,0.025968000292778015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,512,0.012330667012267642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,512,0.024753777517212763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,256,0.011897777517636618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,128,0.002967111145456632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,256,0.023367111881573994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,32,0.012528888881206512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,128,0.01182577759027481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,64,0.0028497777465316984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,4096,128,0.022952889402707417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,4096,64,0.011918221910794577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,4096,32,0.0029395555870400537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,65536,0.16852888796064588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,65536,0.12713689274258083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,16384,0.045724444919162326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,16384,0.040609776973724365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,12288,0.036222222778532244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,12288,0.03458933366669549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,16384,0.05832978089650472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,10240,0.031153778235117596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,10240,0.030584000878863867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,12288,0.05099911159939236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,8192,0.026134222745895386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,7168,0.023753777146339417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,8192,0.026649778087933857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,10240,0.04754844307899475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,65536,0.15935199790530735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,8192,0.0421635541650984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,7168,0.04050755500793457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,7168,0.024375110864639282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,6144,0.018852444158660043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,6144,0.02293511066171858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,5120,0.01586933268441094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,4096,0.019183110859658983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,6144,0.03797333439191183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,3584,0.011884444289737277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,5120,0.020787555310461257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,4096,0.012963555753231049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,5120,0.039464887645509504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,4096,0.035240888595581055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,3584,0.01863555610179901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,3072,0.01074222226937612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,3584,0.03361599975162082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,3072,0.01789422167672051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,2560,0.009832888841629028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,2560,0.017557332913080852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,3072,0.03269066744380527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,1536,0.014402666025691561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,2048,0.00832177781396442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,2560,0.03321866525544061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,1024,0.01329422245422999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,1536,0.03039466672473484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,768,0.0046897778908411665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,2048,0.016172443826993305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,1536,0.007147555549939473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,2048,0.032010667853885226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,1024,0.005673777725961473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,768,0.013017777767446307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,1024,0.026388444834285315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,512,0.024580443898836773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,512,0.0035502223504914176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,128,0.002946666752298673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,768,0.026505778233210247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,512,0.012431999875439538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,256,0.003156444471743372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,64,0.011938666303952536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,256,0.011887110769748688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,128,0.011912888950771756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,256,0.02332266668478648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,64,0.0028320000403457214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3584,32,0.0028853333658642242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3584,128,0.02276800076166789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3584,32,0.012344888514942594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,65536,0.14093955357869467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,65536,0.12284533182779948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,16384,0.041591111156675555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,16384,0.05544355511665344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,16384,0.03937955697377523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,12288,0.03280444608794318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,12288,0.03278933299912347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,10240,0.028245333168241713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,10240,0.029595557186338637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,12288,0.053120000494851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,8192,0.02437244521247016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,10240,0.04497244291835361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,65536,0.15057422055138484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,8192,0.02571022179391649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,7168,0.02199288871553209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,8192,0.04005511270629035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,7168,0.023400889502631292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,6144,0.0367253323396047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,6144,0.019681778219011094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,6144,0.021751110752423603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,7168,0.04191911220550537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,5120,0.017097777790493436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,5120,0.020192000601026747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,3584,0.018262222409248352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,4096,0.011771555576059552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,3072,0.009874666730562845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,5120,0.03652355406019423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,4096,0.018928888771269057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,3584,0.010879110958841113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,3072,0.03241599930657281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,2560,0.01590222285853492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,4096,0.03409511182043288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,3072,0.017553778158293832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,2560,0.008873778084913889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,3584,0.03353777858946059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,2048,0.007710222568776872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,2560,0.03312622176276313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,2048,0.015472888946533203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,1536,0.006728888799746831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,2048,0.031890667147106595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,1536,0.013835555149449242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,1024,0.005141333159473208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,768,0.012876444392734103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,1536,0.02926666537920634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,1024,0.013226666384273104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,768,0.025864000121752422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,512,0.012234666281276278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,768,0.004182222402758068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,512,0.024519110719362896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,128,0.0029377777957253983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,256,0.02316088808907403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,1024,0.026397332549095154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,512,0.0035866668654812705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,256,0.003162666741344664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,256,0.011889778077602386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,128,0.011726222104496427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,64,0.002765333279967308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,3072,128,0.022794667217466567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,65536,0.11973067124684651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,64,0.01187999960449007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,3072,32,0.002867555452717675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,3072,32,0.012308444413873883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,65536,0.12030043866899277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,12288,0.02774133284886678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,16384,0.033971554703182645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,16384,0.038140445947647095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,12288,0.03177777926127116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,10240,0.0285173323419359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,12288,0.04677866564856636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,10240,0.044046223163604736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,16384,0.05355466736687554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,10240,0.023980443676312763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,8192,0.03881511092185974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,7168,0.022628444764349196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,8192,0.020263110597928364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,6144,0.02212000058756934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,65536,0.14256266752878824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,5120,0.014371555712487964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,8192,0.024534222152498033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,7168,0.018381334013409086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,6144,0.0163137780295478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,5120,0.03614133265283372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,4096,0.018698667486508686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,7168,0.03792355457941691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,5120,0.020418667131000094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,6144,0.0359280010064443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,4096,0.012578667038016848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,3584,0.011716444459226398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,4096,0.03446133269204034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,3584,0.017900443739361234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,3072,0.008942222429646386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,3584,0.03305866652064853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,3072,0.017437333861986797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,2560,0.008171555896600088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,3072,0.031786666976081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,2560,0.016531555189026725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,2048,0.007189333438873291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,2560,0.03267022305064731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,2048,0.015056000815497505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,1536,0.0063493334584765965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,2048,0.031164444155163232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,1536,0.013759111364682516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,1024,0.004550222307443619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,1536,0.028549333413441975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,1024,0.013072888884279462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,768,0.003933333274390963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,256,0.0031635556370019913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,1024,0.026366222235891554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,512,0.024438222249348957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,768,0.012912888493802814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,512,0.0035173334181308746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,256,0.023119999302758112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,768,0.025964443882306416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,512,0.012392000191741519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,256,0.01198933356338077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,128,0.0029004443850782183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,128,0.01185066666867998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,64,0.0027724444452259275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2560,128,0.02299733294381036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,64,0.011663999822404651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,16384,0.028677334388097126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2560,32,0.0028151110228565005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2560,32,0.012349333200189801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,65536,0.09668888648351033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,65536,0.11327377955118816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,16384,0.03641155693266127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,12288,0.023158222436904907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,12288,0.04608800013860067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,16384,0.0518124434683058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,12288,0.03036622206370036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,10240,0.02038666605949402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,10240,0.027316444449954565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,65536,0.1401431163152059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,8192,0.016806221670574613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,8192,0.023450666003757056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,10240,0.04492533206939697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,7168,0.01535999940501319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,6144,0.02090666691462199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,8192,0.03750133183267381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,7168,0.022297778063350256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,5120,0.019665777683258057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,6144,0.013964444398880005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,7168,0.03811288873354594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,5120,0.01257955531279246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,6144,0.03570400012864007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,4096,0.011128889189826118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,5120,0.03578844335344102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,4096,0.018350222044520907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,3584,0.010324444207880232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,3072,0.016809778081046212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,4096,0.03366400135887994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,3584,0.017471111483044095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,3072,0.00853066643079122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,3584,0.03319999906751845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,2560,0.00758844448460473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,3072,0.03194933467441135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,2560,0.015055111712879606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,2048,0.006806222101052602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,1536,0.01369599997997284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,2560,0.03073422114054362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,2048,0.014362666341993542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,1536,0.00564444437623024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,2048,0.03001777662171258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,1024,0.004093333250946469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,1024,0.013051555388503604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,1536,0.028166221247778997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,768,0.025695110360781353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,256,0.0030951110853089225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,768,0.003768889026509391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,1024,0.02626222206486596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,768,0.012780444489585029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,512,0.00342399999499321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,512,0.012286221815480126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,512,0.024392889605628118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,256,0.011904888682895236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,128,0.022853333089086745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,128,0.0029031110720502008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,2048,256,0.023021333747439917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,128,0.011725333001878528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,64,0.0027057776848475137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,64,0.011766222616036734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,2048,32,0.0027511111564106415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,2048,32,0.012240000069141388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,65536,0.07378577523761325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,65536,0.10600799984402126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,16384,0.02308533257908291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,16384,0.034648001194000244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,12288,0.01906044450071123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,10240,0.025695110360781353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,12288,0.0288355549176534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,16384,0.04901511139339871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,10240,0.041144887606302895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,65536,0.13026311662462023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,10240,0.016765332884258695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,12288,0.04367466767628988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,8192,0.014105778601434497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,8192,0.02260711126857334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,6144,0.020351111888885498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,7168,0.013013333082199097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,8192,0.03702577617433336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,7168,0.02162666618824005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,6144,0.012048888537618848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,7168,0.03654933306905959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,5120,0.011005333728260465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,6144,0.03490399983194139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,5120,0.019299555155966017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,4096,0.009119999905427298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,5120,0.0351200004418691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,4096,0.01705333259370592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,3584,0.008488000267081791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,4096,0.03361599975162082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,3584,0.015930665863884818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,3072,0.007954667011896769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,3584,0.032123555739720665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,3072,0.015832000308566622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,2560,0.007183999651008182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,3072,0.029021332661310833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,2560,0.014476444986131458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,2048,0.005946666830115848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,2560,0.029623111089070637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,2048,0.01408622165520986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,1536,0.005076444397370021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,2048,0.028811554114023846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,1536,0.01350311107105679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,768,0.012659555508030785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,1024,0.004090666770935059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,1536,0.028212444649802312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,1024,0.012757333616415659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,768,0.0037751110891501107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,1024,0.026107556290096704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,512,0.0034533333447244433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,512,0.012307555311255984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,768,0.025348444779713947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,256,0.0031102223114834893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,256,0.01181688904762268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,512,0.02453777856296963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,128,0.002879111096262932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,256,0.023010666171709698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,128,0.011759111450778114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1536,128,0.02277333372169071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,64,0.0027137777457634607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1536,32,0.002749333365095986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,64,0.011584888729784222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1536,32,0.012156444291273752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,65536,0.05400355656941732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,65536,0.07425510883331299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,16384,0.01722666621208191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,16384,0.026017778449588354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,65536,0.13714310857984754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,12288,0.0140346669488483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,16384,0.04755466514163547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,12288,0.022639999787012737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,10240,0.012960000170601739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,12288,0.0429004430770874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,10240,0.02048711147573259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,8192,0.011212444139851464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,10240,0.041466666592492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,8192,0.019075555933846366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,7168,0.010709332923094431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,8192,0.03673600157101949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,7168,0.018413333429230582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,6144,0.009780444204807281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,6144,0.03473866648144192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,5120,0.016026667422718473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,6144,0.017666666044129264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,7168,0.03640710976388719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,5120,0.00983911090426975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,4096,0.009045333498054082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,5120,0.034824887911478676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,4096,0.015564444992277356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,3584,0.008652444514963362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,3584,0.014640889234013028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,3072,0.02892000145382351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,4096,0.030768000417285498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,3072,0.0075937774446275495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,3584,0.030212445391549006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,3072,0.014447111222479077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,2560,0.006743999818960826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,2560,0.014056889547242058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,1536,0.013218666944238873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,2048,0.0059573331640826324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,2048,0.013656888571050433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,2560,0.029567112525304157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,1536,0.0047786666287316214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,2048,0.02817866537306044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,1024,0.004084444294373195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,1536,0.027726221415731642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,1024,0.012763555679056378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,768,0.003744889050722122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,1024,0.025939555631743536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,768,0.012525333298577202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,512,0.0034222222036785553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,768,0.025504888759719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,512,0.012079999678664737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,256,0.003093333293994268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,512,0.02426044477356805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,256,0.01183555523554484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,128,0.0028684443483750024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,256,0.02311022248533037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,128,0.011601777540312873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,64,0.0027111110587914786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,1024,128,0.022654222117529974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,1024,32,0.0027146666414207886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,64,0.011526222030321756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,1024,32,0.01180622229973475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,65536,0.044080889887279935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,65536,0.12061956193712021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,12288,0.02233155568440755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,10240,0.011271110839313932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,65536,0.07008533345328437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,16384,0.014461333552996317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,16384,0.02508977717823452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,12288,0.012215110990736218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,16384,0.047361777888404004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,12288,0.04271822174390157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,10240,0.020491555333137512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,8192,0.011359111302428775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,10240,0.040115555127461754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,8192,0.018981332580248516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,7168,0.010661332971519895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,8192,0.036377777655919395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,7168,0.017258667283587985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,6144,0.010015111002657149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,6144,0.01664266652531094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,7168,0.03613333238495721
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,5120,0.009295110901196798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,6144,0.03332355618476868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,5120,0.015508444772826301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,4096,0.008346666892369589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,5120,0.032728887266582914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,4096,0.01495822270711263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,3584,0.00756888919406467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,4096,0.03101955519782172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,3584,0.014646222194035849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,3072,0.006776000062624614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,3584,0.02957688768704732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,3072,0.01428533262676663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,2560,0.0063946665161185795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,2560,0.013903111219406128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,3072,0.028760890165964764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,2048,0.005415110952324337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,2048,0.013442666994200813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,2560,0.02905333373281691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,1536,0.0047102222839991255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,1536,0.013047110703256396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,2048,0.02817866537306044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,1024,0.0040702223777771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,1536,0.027250667413075764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,1024,0.012657778130637275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,768,0.003685333248641756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,1024,0.026001777913835313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,768,0.012411555482281579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,512,0.0033555556502607134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,768,0.02535022298494975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,512,0.011923555698659686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,256,0.003022222262289789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,512,0.02431022293037838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,256,0.011791110866599612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,128,0.0028071111689011254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,256,0.02285066743691762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,128,0.011545778148704104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,32,0.01180622229973475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,64,0.0026942222482628296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,768,128,0.022614222433831956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,768,64,0.011427555647161273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,768,32,0.0027013334135214486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,65536,0.032813333802753024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,65536,0.0666560000843472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,16384,0.011344888971911537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,16384,0.024831111232439678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,65536,0.12114932801988389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,12288,0.011357333097192975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,12288,0.022097777989175584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,16384,0.04691911074850294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,10240,0.010416888528399997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,12288,0.04270311196645101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,10240,0.019959110352728102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,8192,0.009491555392742157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,8192,0.01735111077626546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,10240,0.03994044330384996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,7168,0.008994667066468133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,8192,0.034576889541414045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,7168,0.016549333930015564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,6144,0.008284444610277811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,6144,0.015765332513385348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,7168,0.03408888975779215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,5120,0.007531555162535773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,6144,0.03229511115286086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,5120,0.015023110641373528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,4096,0.006519111080302133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,5120,0.0318195554945204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,4096,0.014555555250909595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,3584,0.0068711112770769335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,2560,0.005964444329341252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,4096,0.0307004451751709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,3584,0.014142221874660917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,3072,0.006592000110281839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,3584,0.029419554604424372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,3072,0.013812444276279874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,3072,0.028255999088287354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,2560,0.013439111411571503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,2048,0.0053137776752312975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,2560,0.029042667812771265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,2048,0.013156444662147097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,1536,0.004642666627963384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,2048,0.027814222706688776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,1536,0.012867555850081973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,1024,0.003913777983850903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,1536,0.02732266651259528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,1024,0.012496888637542725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,768,0.0035946667194366455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,1024,0.02548800077703264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,768,0.012225777738624148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,512,0.0033048888047536216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,768,0.02509511179394192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,512,0.011824000212881299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,256,0.003006222140457895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,512,0.02418488926357693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,256,0.011590222517649332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,128,0.0027511111564106415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,256,0.022870222727457683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,32,0.01162577751610014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,128,0.011387555963463254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,64,0.00258933338854048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,512,128,0.022593778040673997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,512,64,0.011398221883508893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,512,32,0.0026231110095977783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,65536,0.022288888692855835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,65536,0.06431288851632012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,16384,0.009753777749008602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,65536,0.12211467160118951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,16384,0.024005333582560223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,10240,0.01828799976242913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,12288,0.008819555242856344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,16384,0.04646044307284885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,12288,0.020471110939979553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,10240,0.007710222568776872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,12288,0.04031822085380554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,8192,0.007044444481531779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,10240,0.03944622145758735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,8192,0.017066667477289837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,7168,0.006690666493442323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,6144,0.015602666470739575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,8192,0.03394133183691237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,6144,0.03195200032658047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,7168,0.016342222690582275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,6144,0.006344888773229387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,7168,0.0337084432442983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,5120,0.006825777805513806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,5120,0.014999111493428549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,4096,0.006455111006895701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,3584,0.014152889450391134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,4096,0.030580444468392268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,4096,0.014468444718254937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,3072,0.013839999834696451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,5120,0.03144800000720554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,3584,0.006883555402358373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,3072,0.006623111251327727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,3584,0.02928000026279026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,2560,0.005958222266700532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,2560,0.013540444274743399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,3072,0.02829688787460327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,2048,0.0053075556125905775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,1536,0.01290933373901579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,2048,0.013187555803192986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,2560,0.02875022093454997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,1536,0.004621333132187526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,2048,0.02803200152185228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,1024,0.00397866674595409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,1536,0.027341332700517442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,1024,0.012454222473833295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,768,0.003605333467324575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,1024,0.025614221890767414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,768,0.01221333361334271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,512,0.0032586666444937387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,512,0.01201244443655014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,768,0.025197333759731714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,256,0.022966222630606756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,256,0.002926222152180142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,512,0.024007999234729346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,256,0.011722666521867117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,128,0.002768888862596618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,128,0.011355555719799466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,64,0.0026062221990691293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,64,0.011432888607184092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,256,32,0.002604444407754474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,256,128,0.02275733318593767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,65536,0.013750221994188098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,256,32,0.011635555161370171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,16384,0.04432266619470385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,65536,0.06528800063663058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,16384,0.006681777950790193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,16384,0.02272622287273407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,65536,0.12216356065538193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,12288,0.006550222221348021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,12288,0.01976977785428365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,10240,0.006722666737106111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,10240,0.01830044388771057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,12288,0.04020533296797011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,7168,0.016373333003785875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,8192,0.03392533461252848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,8192,0.006680000159475539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,10240,0.03741777936617533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,8192,0.0169493340783649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,7168,0.006515555497672822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,6144,0.006346666564544042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,6144,0.015626667274369132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,7168,0.03362577822473314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,4096,0.014376888672510782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,5120,0.00675022229552269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,6144,0.03189244535234239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,4096,0.030431111653645832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,5120,0.014911111858155994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,4096,0.00637155564294921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,5120,0.03168355425198873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,2560,0.005922666854328579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,3584,0.006770666688680649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,3072,0.028348445892333984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,3584,0.01401422255569034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,3072,0.0064586665895250105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,3584,0.02951999836497837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,3072,0.013830222189426422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,2560,0.013453333742088742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,1536,0.012912888493802814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,2048,0.005261333452330695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,2048,0.013119999733236102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,2560,0.028677334388097126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,1536,0.004604444321658877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,2048,0.027838223510318335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,1024,0.003941333128346337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,1536,0.027150221996837195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,1024,0.012442666623327466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,512,0.011887110769748688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,768,0.003573333223660787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,1024,0.02535199953450097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,768,0.012272000312805176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,512,0.003273777870668305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,768,0.024970667229758367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,256,0.022774222824308608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,128,0.01148088855875863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,256,0.002944888960984018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,512,0.02381955583890279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,256,0.011730666789743634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,128,0.002757333426011933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,64,0.002609777781698439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,64,0.011249778171380361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,48,128,128,0.02257777750492096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,128,32,0.0026444445053736367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,128,32,0.011601777540312873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,65536,0.012451555165979596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,65536,0.06444088617960612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,16384,0.006493333313200209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,16384,0.022522666388087805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,12288,0.0063999998900625445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,12288,0.019778667224778067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,10240,0.00657155571712388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,8192,0.006432000133726332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,10240,0.018178666631380718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,8192,0.016955556141005624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,7168,0.006314666734801398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,7168,0.01628266606065962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,6144,0.006287999865081575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,6144,0.015590222345458137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,3072,0.006423111177153057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,5120,0.006657777975002925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,5120,0.014888000157144336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,4096,0.006224000205596288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,2560,0.013343111508422427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,3584,0.006714666469229593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,4096,0.014344000154071383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,3584,0.014007111390431723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,3072,0.013717333475748697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,2560,0.005927111125654644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,1024,0.012332444389661154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,2048,0.005241777747869492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,1536,0.004586666822433472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,2048,0.01316088851955202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,1536,0.012796444197495779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,1024,0.00388711111413108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,128,0.002736888825893402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,768,0.003567111161020067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,64,0.0025804444319672054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,512,0.003208000005947219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,64,0.011300444602966309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,768,0.012186666329701742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,512,0.011963555382357704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,256,0.0029013332807355454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,256,0.01163822246922387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,128,0.01146755533085929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,64,32,0.002607999990383784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,64,32,0.011517333487669626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,65536,0.011622221933470832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,65536,0.0638622244199117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,16384,0.006450666735569636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,16384,0.02251822253068288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,12288,0.006317333214812809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,12288,0.01971733404530419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,10240,0.006505777852402792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,8192,0.006439111298984951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,10240,0.018301332990328472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,8192,0.017072000437312655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,7168,0.006291555447710886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,6144,0.006224888894293044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,4096,0.014335999886194864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,7168,0.016233777006467182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,5120,0.006613333192136552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,6144,0.015603555573357476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,3072,0.013624000052611033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,5120,0.015053333507643806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,4096,0.006243555496136348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,3584,0.006634666687912411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,3584,0.014081777797804939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,3072,0.00638755535085996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,2560,0.0058808889653947614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,2560,0.013396444420019785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,2048,0.005229333208666907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,2048,0.013103110922707452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,768,0.012143111063374413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,1536,0.00453333349691497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,1536,0.012744888663291931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,1024,0.0039013334446483185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,1024,0.012455110748608908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,768,0.003543111185232798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,512,0.0032222221295038858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,512,0.011950222154458364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,256,0.0028826666788922418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,256,0.011539555258221097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,32,0.011501332951916588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,128,0.0026906666656335196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,128,0.011456888582971362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,64,0.002554666664865282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,48,32,64,0.011251555548773872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,48,32,32,0.0025555555605226094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,16384,0.37073598967658145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,12288,0.2869955433739556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,16384,0.7141831186082629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,12288,0.5441768964131674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,10240,0.4543306562635634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,10240,0.24638133578830293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,8192,0.20481244723002115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,8192,0.3660417662726508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,7168,0.3168497880299886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,7168,0.19536266061994764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,6144,0.15594844023386636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,6144,0.2731155554453532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,5120,0.2696604463789198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,5120,0.1299146678712633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,8192,0.23867289225260416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,4096,0.18513333797454834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,10240,0.2889137797885471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,4096,0.1069271100891961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,12288,0.3427893320719401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,3584,0.16207555929819742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,6144,0.1812302271525065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,7168,0.21322666274176705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,3584,0.09684711032443577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,3072,0.14008622699313694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,3072,0.08609777688980103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,5120,0.16110755337609187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,2560,0.11699733469221328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,2560,0.08040177822113037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,2048,0.09346755345662434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,16384,0.4051324526468913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,2048,0.06350844436221652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,1536,0.07262577613194783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,4096,0.13419289059109157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,1536,0.053242666853798755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,1024,0.04939377639028761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,3584,0.11289332972632514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,1024,0.0424924459722307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,2560,0.09594488806194729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,2048,0.07835733228259616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,768,0.04499288731151157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,768,0.03844533363978068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,1536,0.06615555286407471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,3072,0.11611378192901611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,512,0.02756888833310869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,1024,0.05432266659206814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,512,0.0348693331082662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,768,0.04717422193951077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,256,0.015758223003811307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,256,0.028619554307725694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,512,0.04080444574356079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,128,0.010460444622569613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,256,0.03553422292073568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,128,0.0270480000310474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,64,0.006899555524190267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,65536,128,0.029847111966874864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,64,0.025124443901909724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,65536,32,0.005162666655249066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,65536,32,0.024615110622511968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,16384,0.17609333992004395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,16384,0.1080568896399604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,12288,0.13419822851816812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,65536,0.4353920088873969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,12288,0.0863484475347731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,65536,0.7037368880377876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,10240,0.11131110456254746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,10240,0.07387289073732164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,8192,0.09047910902235244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,8192,0.061569776799943715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,7168,0.07956977685292561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,7168,0.05551022291183472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,6144,0.06950310866038005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,12288,0.10139999787012736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,16384,0.12093688382042779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,6144,0.050168000989490084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,8192,0.07497688796785143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,5120,0.05819733275307549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,5120,0.044432888428370156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,7168,0.07764977878994413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,4096,0.04754844307899475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,4096,0.03817866577042474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,6144,0.07051377826266818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,10240,0.15151200029585096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,3584,0.04194133480389913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,3584,0.03637511200375027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,5120,0.05888266695870293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,3072,0.03669688767857022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,4096,0.05264800124698215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,3072,0.0555573304494222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,2560,0.03110755483309428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,3584,0.05119911167356703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,2560,0.030525333351559106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,3072,0.048193779256608754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,2048,0.025701334079106648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,2048,0.03190044562021891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,2560,0.04521066612667508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,1536,0.020245333512624104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,1536,0.024015110399987962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,2048,0.04123288724157546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,1024,0.014306666122542487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,1536,0.03661155700683594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,1024,0.022890667120615642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,768,0.011287111375066968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,1024,0.03310666812790765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,768,0.020056888461112976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,512,0.008580444587601556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,768,0.03102488981352912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,512,0.018759111563364666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,256,0.005771555420425203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,512,0.02888266576661004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,256,0.013911111487282647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,128,0.0036311112344264984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,128,0.023341332872708637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,64,0.014754666222466363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,256,0.024171556035677593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,32,0.014671999547216626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,16384,128,0.014863111906581454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,64,0.0033448889023727844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,16384,65536,0.4240844514634874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,16384,32,0.0034168888297345904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,65536,0.31372088856167263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,16384,0.14012622833251953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,16384,0.09122221999698216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,65536,0.5456417931450738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,12288,0.10977688762876724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,12288,0.07284088929494222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,10240,0.09151022301779853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,10240,0.0634746683968438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,8192,0.07467910978529188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,8192,0.05297422077920702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,7168,0.06528355677922566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,16384,0.10542044374677871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,7168,0.04904799991183811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,10240,0.07517421907848783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,6144,0.05657866928312513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,6144,0.04438933398988512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,5120,0.04731377628114489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,8192,0.07076888614230685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,12288,0.13730666372511122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,5120,0.03808266586727566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,7168,0.06167733007007175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,4096,0.038951999611324735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,6144,0.05728622277577718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,4096,0.03439022103945414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,3584,0.03435644507408142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,5120,0.052652445104387074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,3072,0.02903199858135647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,3584,0.0314995547135671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,3072,0.02997777859369914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,4096,0.07190399699740939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,3584,0.046443555090162486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,2560,0.025595555702845257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,2560,0.027126222848892212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,3072,0.043861332866880626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,2048,0.021163556310865615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,2560,0.041005333264668785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,2048,0.028912001185946997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,1536,0.01664355562792884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,1536,0.03420977791150411
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,1024,0.019842666056421068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,1536,0.02176977859603034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,2048,0.03735111157099406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,1024,0.011901333100265928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,768,0.009619555539555019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,768,0.018617777360810172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,256,0.004495999879307217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,1024,0.03461600012249417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,65536,0.36571823226080996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,512,0.007357333269384172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,768,0.030020445585250854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,128,0.013832000394662222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,512,0.01640977793269687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,512,0.028352889749738906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,256,0.01368977791733212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,128,0.003698666476541095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,256,0.023735110958417256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,64,0.0032257777121331957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,12288,32,0.0031822222388452957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,64,0.014183999763594734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,12288,128,0.023179555932680767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,12288,32,0.01407911048995124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,16384,0.11412710613674587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,65536,0.2717546621958415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,16384,0.07767199807696872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,65536,0.4551973342895508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,12288,0.08742488755120172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,12288,0.06255822049246894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,10240,0.07633866866429646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,10240,0.05456977751519945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,8192,0.05929599867926704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,16384,0.09196444352467854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,8192,0.046661333905326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,12288,0.07676444451014201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,7168,0.052221334642834134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,7168,0.042824887567096286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,10240,0.07407822211583455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,6144,0.04621866676542494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,6144,0.03904177745183309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,8192,0.09121155738830566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,7168,0.057674666245778404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,5120,0.03872088922394647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,5120,0.03494044476085239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,4096,0.031545778115590416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,6144,0.05329599976539612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,4096,0.035924444595972695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,5120,0.051557332277297974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,3584,0.027983112467659846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,4096,0.044534222947226636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,3584,0.029498666524887085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,3072,0.02499288817246755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,3072,0.027071999178992376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,3584,0.0439991123146481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,2560,0.021075555019908484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,65536,0.31587200694613987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,3072,0.04161333375506931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,2560,0.02527911133236355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,2048,0.017440888616773818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,1536,0.02111111084620158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,2560,0.03942755526966519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,2048,0.02273955610063341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,1536,0.013649777405791812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,2048,0.0355751117070516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,1024,0.010062221851613786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,768,0.0177839994430542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,1536,0.033576889170540705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,1024,0.018954666124449838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,768,0.008374222450786168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,1024,0.035232888327704534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,512,0.006859555426571105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,768,0.02983199887805515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,512,0.014741332994567024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,256,0.003855111284388436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,512,0.02530044482813941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,256,0.013660444153679742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,128,0.0032951111594835916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,256,0.024874667326609295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,128,0.013439111411571503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,64,0.003028444531891081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,10240,128,0.02335111134582096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,64,0.013814222481515674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,10240,32,0.003075555587808291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,10240,32,0.01383733335468504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,65536,0.20853422747717965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,16384,0.08941333161460029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,65536,0.3574684460957845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,16384,0.06298044655058119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,12288,0.06864000029034086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,12288,0.051560001240836255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,10240,0.057815108034345836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,10240,0.04486844605869717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,16384,0.07891733116573758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,8192,0.04705244302749634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,8192,0.03886577818128798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,12288,0.08980088763766819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,7168,0.047989332013660006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,10240,0.06131466892030504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,6144,0.03896711270014445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,7168,0.035205334424972534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,8192,0.0543066660563151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,6144,0.03679022192955017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,7168,0.05184177888764275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,5120,0.031014220582114324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,5120,0.03228977653715346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,6144,0.0481333335240682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,4096,0.025511110822359722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,4096,0.026103999879625108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,3072,0.0200284438000785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,65536,0.26794311735365123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,5120,0.04527555571662056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,3584,0.022745778163274128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,3584,0.02494311167134179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,4096,0.04069955481423272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,3584,0.04084622197681003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,3072,0.021817778547604878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,2560,0.017159110969967313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,3072,0.04028977619277106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,2560,0.02033066584004296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,2048,0.014165333575672574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,2560,0.03607822126812405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,2048,0.019173334042231243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,1536,0.011321778098742167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,2048,0.034239110019471906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,1536,0.018413333429230582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,768,0.013423999978436364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,1024,0.008452444440788692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,1536,0.032419555717044406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,1024,0.01623288955953386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,768,0.0072266666425599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,1024,0.029899554120169744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,512,0.0057946667075157166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,512,0.012657778130637275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,768,0.028805332051383123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,128,0.011944000091817645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,256,0.003666666646798452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,512,0.02437777817249298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,256,0.01221777747074763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,128,0.0032417778339650896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,256,0.023532445232073467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,64,0.003108444520168834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,8192,128,0.023022222850057814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,64,0.012115555504957834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,16384,0.07988888687557645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,65536,0.3323475519816081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,65536,0.19247378243340385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,8192,32,0.003112888791494899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,8192,32,0.012048000262843238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,12288,0.048115554783079356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,16384,0.058443552917904325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,12288,0.07149155272377862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,10240,0.05152177810668945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,10240,0.04153600003984239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,16384,0.07673333088556926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,8192,0.04222844375504387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,12288,0.06398133436838786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,8192,0.03629066546758016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,7168,0.03715644280115763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,10240,0.058928887049357094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,7168,0.03344710999064975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,8192,0.05236444539493985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,6144,0.032835556401146784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,5120,0.027875555886162653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,6144,0.030516445636749268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,7168,0.05004000001483493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,5120,0.027965333726671007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,6144,0.046161777443355985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,4096,0.023346667488416035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,65536,0.23985245492723253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,4096,0.024382222029897902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,5120,0.043735110097461276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,3072,0.021149333980348375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,3584,0.02067466742462582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,4096,0.039453334278530545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,3072,0.037245333194732666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,3584,0.02219733264711168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,3072,0.01850488947497474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,2048,0.018439110782411363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,3584,0.03864799936612447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,2560,0.015671110815472074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,2560,0.01977511081430647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,2048,0.013010666602187686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,2560,0.03664888938268026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,1536,0.01054666688044866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,2048,0.03338844577471415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,1024,0.02966044346491496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,1536,0.017272000511487324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,512,0.005502222312821283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,1024,0.00798044436507755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,1536,0.03159555461671617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,1024,0.015234667393896313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,768,0.006787555499209299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,768,0.013050666285885705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,768,0.026700443691677515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,512,0.012587555580668978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,256,0.0033555556502607134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,512,0.024225778049892847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,256,0.011925333076053195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,128,0.003118222165438864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,256,0.02345866627163357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,128,0.012040889097584618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,64,0.002926222152180142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,7168,128,0.023027555810080633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,64,0.012039999994966718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,7168,32,0.0029768887907266617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,7168,32,0.012005333271291522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,65536,0.17658755514356825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,65536,0.2694151136610243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,16384,0.07443910837173462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,16384,0.05542844533920288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,10240,0.05359733435842726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,12288,0.056208001242743603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,12288,0.04548977812131246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,16384,0.07011377811431885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,8192,0.03510133425394694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,10240,0.03938577903641595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,12288,0.05928711096445719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,8192,0.038902223110198975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,10240,0.054923554261525474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,7168,0.03400177756945292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,6144,0.028891556792789038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,7168,0.03150755498144362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,8192,0.04889333248138428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,6144,0.030032889710532293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,7168,0.047316445244683154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,5120,0.02525866693920559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,5120,0.02599999970859951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,6144,0.04442400071356031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,4096,0.021493333909246657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,5120,0.041532443629370794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,4096,0.022655111220147874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,3584,0.019160888261265226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,3072,0.020234665936893888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,3584,0.021474666065639917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,4096,0.03727644350793626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,3072,0.016731555263201397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,2048,0.0122417774465349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,3584,0.037267555793126426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,65536,0.2620320055219862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,2560,0.01462933255566491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,2560,0.019380443625979953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,3072,0.035628444618648954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,2560,0.03446311089727614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,2048,0.017736888594097562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,1536,0.009924444059530893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,2048,0.03277333246337043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,1536,0.021353777911927965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,1024,0.02886577778392368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,1024,0.007495111061467065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,1536,0.03138844503296746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,1024,0.014078223043017916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,512,0.012496888637542725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,768,0.006339555399285422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,768,0.013218666944238873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,512,0.00444088876247406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,768,0.02667288813326094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,256,0.0033742222521040174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,512,0.024422221713595923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,256,0.01200177768866221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,128,0.00309244439833694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,256,0.023238221804300945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,128,0.011779555843936073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,64,0.0029235554652081597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,6144,128,0.02313333253065745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,64,0.011997333003415002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,6144,32,0.002945777856641345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,6144,32,0.01200888885392083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,65536,0.2261751227908664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,65536,0.15419199731614855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,16384,0.05858044491873848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,16384,0.05074310965008206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,12288,0.04520888792143928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,10240,0.038223998414145574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,12288,0.041758222712410815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,16384,0.06725333134333293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,10240,0.036870221296946205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,8192,0.031153778235117596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,12288,0.05772799915737576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,8192,0.03200177682770623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,7168,0.02773422168360816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,10240,0.05386666787995232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,7168,0.029937777254316542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,8192,0.04750222298834059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,65536,0.18233688672383627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,6144,0.024655999408827886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,6144,0.026896889011065166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,7168,0.04737600021892124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,5120,0.021200888686709937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,6144,0.043502221504847206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,5120,0.02346933384736379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,4096,0.03657066822052002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,4096,0.017603556315104168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,4096,0.021264889174037512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,5120,0.04066577884886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,3584,0.015871110889646742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,3584,0.02018311123053233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,3072,0.0136808885468377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,3072,0.01944177846113841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,3584,0.03638666537072923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,2560,0.011961778004964193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,2560,0.01841600073708428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,3072,0.035084442959891424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,2048,0.010178666975763109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,2560,0.03417866759830051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,2048,0.01716800034046173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,1536,0.008457777400811514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,2048,0.032371554109785296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,1536,0.015855110353893705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,1536,0.0313022236029307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,1024,0.006601777755551868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,1024,0.013400889105266996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,768,0.005691555639108022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,1024,0.029077334536446467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,768,0.01279022213485506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,512,0.003579555700222651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,768,0.02528800070285797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,512,0.012411555482281579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,128,0.01167555567291048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,256,0.003236444460021125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,512,0.024122666981485155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,256,0.011995555626021491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,128,0.003014222201373842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,32,0.011910222470760345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,256,0.02346222268210517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,64,0.0028488888508743713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,65536,0.13427110513051352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,16384,0.04973422156439888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,5120,128,0.022966222630606756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,5120,64,0.011796444654464722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,5120,32,0.0028799999919202593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,65536,0.18062844541337755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,16384,0.04424533247947693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,12288,0.03958844476275974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,12288,0.03609600000911289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,16384,0.0619973341623942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,10240,0.03389244609408908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,10240,0.033370667033725314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,12288,0.053836445013682045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,8192,0.02804888950453864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,8192,0.0444453325536516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,8192,0.02903999884923299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,6144,0.022806222240130108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,10240,0.05063910947905647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,7168,0.025431111454963684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,65536,0.16855200131734213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,7168,0.026380444566408794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,6144,0.023955555425749883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,7168,0.04330133398373922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,4096,0.014567999376191033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,5120,0.017664889494578045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,6144,0.039860443936453924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,5120,0.02186044388347202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,5120,0.03805333375930786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,4096,0.01997777819633484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,3584,0.013065777719020844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,3072,0.018191999859280057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,3584,0.019349333312776353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,2560,0.010292444792058732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,4096,0.03516711129082574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,3072,0.011647111011876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,3584,0.03535822365019057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,2560,0.01754755609565311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,1536,0.007347555624114142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,3072,0.03408177693684896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,1536,0.014707555373509726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,2048,0.008820444345474243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,2048,0.01644622286160787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,2560,0.03335377905103896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,2048,0.03165155649185181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,1024,0.006058666855096817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,1024,0.027659555276234944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,768,0.012854222622182635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,1536,0.03019022279315525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,1024,0.013363555901580386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,768,0.004746666798988978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,512,0.0034737777378824023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,256,0.01202222208182017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,512,0.012418666647540199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,768,0.02496266696188185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,256,0.0031617778456873367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,512,0.024130667249361675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,128,0.0028862222615215513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,256,0.02316177719169193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,128,0.011908444265524546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,64,0.0027840000887711844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,4096,128,0.022853333089086745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,64,0.011857777833938599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,4096,32,0.002924444360865487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,4096,32,0.01166311071978675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,65536,0.16552266809675428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,65536,0.1271617809931437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,16384,0.044581333796183266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,16384,0.04223733478122287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,12288,0.03551910983191596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,12288,0.03425333234998915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,16384,0.06007022327846951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,10240,0.03048355711830987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,10240,0.031151112582948472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,12288,0.05170933405558268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,8192,0.025655110677083332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,7168,0.02539377742343479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,8192,0.04376444551679823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,8192,0.02793688906563653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,10240,0.049124442868762545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,7168,0.023173333870040044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,65536,0.15758400493197972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,5120,0.021291555629836187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,6144,0.020756444997257657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,7168,0.042621334393819175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,6144,0.022826666633288067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,5120,0.01812088820669386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,6144,0.0391902228196462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,4096,0.01294844432009591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,4096,0.01940355532699161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,5120,0.037400000625186496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,3584,0.011844444606039258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,4096,0.03511733478969998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,3072,0.0340320004357232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,3584,0.018792000081804063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,3072,0.010655110908879174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,3584,0.03525955478350321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,3072,0.01849244369400872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,2560,0.009486222432719337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,2560,0.017085333665211994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,2048,0.008048888709810045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,2560,0.03318488929006789
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,2048,0.01629866659641266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,1536,0.029510223203235205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,1536,0.006980444822046492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,2048,0.03145066565937466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,1536,0.015219555960761176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,1024,0.0058373332851462895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,1024,0.013212444053755866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,768,0.004392000122202767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,1024,0.026103110777007207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,768,0.012539555629094442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,512,0.003516444315512975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,768,0.025067556235525344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,512,0.012441777520709567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,256,0.003096888876623578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,512,0.023967999551031325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,256,0.011892444557613797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,32,0.0027831111931138565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,64,0.011762667033407422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,128,0.0028533333291610083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,256,0.02334311107794444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,128,0.011834666960769229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3584,64,0.0028106667515304354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3584,128,0.022710222336981032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3584,32,0.011746666497654386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,65536,0.13911199569702148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,65536,0.12343466281890869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,16384,0.04139022363556756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,12288,0.033266666862699724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,16384,0.0402800010310279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,12288,0.03276266654332479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,16384,0.05949422385957506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,10240,0.027922666735119287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,12288,0.05093244380421109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,10240,0.029640889830059473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,8192,0.023972445064120825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,10240,0.048511998520957104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,8192,0.0270471109284295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,7168,0.021546666820844013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,7168,0.02422844370206197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,65536,0.18614755736456978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,8192,0.04273600048489041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,6144,0.01919644408755832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,7168,0.041315555572509766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,6144,0.02234577801492479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,5120,0.016844444804721408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,6144,0.03899377915594313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,5120,0.02184266679816776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,4096,0.01463644372092353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,3584,0.018210666047202218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,5120,0.0373946660094791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,4096,0.019159111711714003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,3584,0.010778666370444827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,4096,0.03482133481237624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,3072,0.009843555589516958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,3072,0.01757688820362091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,3584,0.0355004436439938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,2048,0.015783111254374187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,2560,0.00869599978129069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,3072,0.03346666693687439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,1536,0.014008000493049622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,2560,0.016827555166350473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,2048,0.007589333587222629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,2560,0.032568001084857516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,1536,0.006535999890830781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,2048,0.0314720008108351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,768,0.012665777570671506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,1024,0.005658666706747479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,1536,0.028020444843504164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,512,0.012392889294359418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,1024,0.01330844478474723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,768,0.004108444270160464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,1024,0.02591466738118066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,128,0.0028728888266616394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,512,0.0035457776652442086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,768,0.02531822191344367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,64,0.002736888825893402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,256,0.0031591111587153543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,512,0.023757333556811016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,256,0.011898666620254517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,32,0.01165688865714603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,256,0.023141332798533972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,128,0.01184088902340995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,65536,0.11974222130245632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,3072,128,0.022562666071785822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,3072,32,0.0027831111931138565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,3072,64,0.011731555892361535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,65536,0.11836889055040146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,12288,0.03184977836079068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,16384,0.05630399783452352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,16384,0.03380889031622145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,16384,0.03868177864286635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,12288,0.027181333965725366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,10240,0.023562666442659166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,8192,0.02481866710715824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,12288,0.04916888806555006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,10240,0.028415110376146104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,8192,0.020040000478426617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,65536,0.14262489477793375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,10240,0.046534223688973315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,6144,0.022062222162882488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,7168,0.018062222335073683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,8192,0.041859553919898145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,7168,0.02327199942535824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,6144,0.016149333781666227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,7168,0.041738665766186185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,5120,0.036646223730511136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,4096,0.018983999888102215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,5120,0.01423555612564087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,6144,0.0378479990694258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,5120,0.020512888828913372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,4096,0.012335999972290464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,3584,0.011535999675591787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,4096,0.03416622347301907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,3584,0.0180711117055681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,3072,0.01073333372672399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,3584,0.03418133325046963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,3072,0.017439110411538016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,2560,0.008220444122950235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,3072,0.033152888218561806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,2560,0.016528889536857605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,2048,0.007086222370465596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,2560,0.032229334115982056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,2048,0.014895111322402954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,1536,0.006176888942718506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,1024,0.013249778085284762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,2048,0.029192888074451025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,1536,0.014210666219393412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,1024,0.004567999806669024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,1536,0.02811466654141744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,768,0.0038480001191298165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,1024,0.025831111603313025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,768,0.0125946667459276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,512,0.0034844444857703317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,256,0.011935110721323224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,768,0.024958221448792353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,512,0.012173333101802401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,256,0.02306755549377865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,128,0.011749332977665795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,256,0.0031306667046414483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,512,0.02398666739463806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,128,0.0028817777832349143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,64,0.00272444449365139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2560,128,0.022699554761250813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,64,0.011605333122942181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,65536,0.11161422729492188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2560,32,0.0027724444452259275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2560,32,0.01147911118136512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,65536,0.09394577476713394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,16384,0.027875555886162653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,12288,0.029913776450686987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,16384,0.0554711123307546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,16384,0.03648266527387831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,12288,0.022687999738587275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,10240,0.019831111033757527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,10240,0.026611554953787062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,12288,0.0484862228234609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,65536,0.14546310901641846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,7168,0.022568888134426538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,8192,0.01644444465637207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,10240,0.045328001181284584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,8192,0.02388711108101739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,7168,0.015023110641373528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,6144,0.013678222066826291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,8192,0.04064177804523044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,7168,0.040172444449530706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,6144,0.02127911150455475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,5120,0.012400888734393649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,5120,0.019866666860050626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,6144,0.03789600067668491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,4096,0.010947555303573608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,4096,0.018534221582942538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,5120,0.03571022219128079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,3584,0.010303110712104374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,4096,0.03400266501638625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,3584,0.01791200041770935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,3072,0.00941422250535753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,2048,0.006730666591061487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,3584,0.03496888942188687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,3072,0.015965332587560017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,2560,0.00742222203148736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,3072,0.03245866629812453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,2560,0.014979556202888489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,2048,0.014264000786675347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,2560,0.032080001301235624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,1024,0.013047999805874295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,1536,0.005438222239414851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,2048,0.02879200047916836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,1536,0.013565333353148567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,1024,0.004063111212518481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,1536,0.027199111051029626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,768,0.00371111101574368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,1024,0.02576622201336755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,768,0.012607110871209038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,512,0.023723555935753718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,512,0.003411555455790626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,768,0.024837333295080397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,512,0.012084444363911947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,256,0.003043555551105075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,256,0.011977777712874942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,64,0.011518222590287527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,128,0.0028364445186323593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,256,0.02309866746266683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,128,0.011786667009194693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,64,0.0026684444811609057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,2048,32,0.002716444432735443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,2048,128,0.02292177743381924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,2048,32,0.01148000028398302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,65536,0.07306400272581312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,65536,0.10296089119381374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,16384,0.022859555151727464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,16384,0.03433422247568766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,12288,0.018527110417683918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,16384,0.05346044566896227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,65536,0.13596622149149576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,12288,0.028732445504930284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,10240,0.01628888812330034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,7168,0.013243555194801755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,12288,0.04673244555791219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,10240,0.025431111454963684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,8192,0.013849777479966482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,10240,0.04407555527157254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,8192,0.022944889134830896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,8192,0.040334221389558576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,7168,0.021785777476098802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,6144,0.011926222178671094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,7168,0.0395706660217709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,6144,0.020632889535692003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,4096,0.00978133330742518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,5120,0.010956444674068026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,6144,0.037272890408833824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,5120,0.019188443819681805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,4096,0.017786666750907898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,5120,0.03611466619703505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,3584,0.009366222553782994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,4096,0.03351555599106683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,3584,0.016392888294325936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,3072,0.008823110825485652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,2560,0.014677332507239448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,3584,0.03286577926741706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,3072,0.015025777949227227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,2560,0.007119110888904995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,1536,0.004956444518433677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,3072,0.03156355685657925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,2048,0.0058959999846087555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,2560,0.029447111818525527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,2048,0.014128888646761576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,2048,0.028493334849675495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,1536,0.013447110851605734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,1024,0.004051555362012652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,1536,0.02695377833313412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,1024,0.012774222426944308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,768,0.024878222081396315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,256,0.002986666642957263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,768,0.003714666598372989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,1024,0.025614221890767414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,768,0.012614222036467658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,512,0.0033573332346147964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,512,0.012166221936543783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,512,0.023642667465739783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,256,0.011772444678677453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,32,0.002699555622206794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,128,0.002809777855873108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,256,0.023016000787417095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,128,0.011734222372372946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1536,64,0.0027022223091787765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1536,128,0.022629333866967097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,64,0.01146399974822998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1536,32,0.011407111254003314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,65536,0.05242844422658285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,65536,0.06477244695027669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,12288,0.02086755633354187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,16384,0.017412443955739338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,16384,0.02368800010946062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,12288,0.014087110757827759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,16384,0.04517244299252828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,65536,0.1171297762129042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,10240,0.012627555264366997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,12288,0.04121422105365329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,10240,0.019744889603720773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,8192,0.011051555474599203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,10240,0.039629333549075656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,8192,0.018759111563364666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,6144,0.017101332545280457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,7168,0.010448888772063784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,8192,0.036547554863823786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,7168,0.017885333961910672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,6144,0.009257777697510189
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,7168,0.03692533241377936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,4096,0.01499111122555203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,5120,0.009105777574910058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,5120,0.015257777439223396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,6144,0.034662223524517484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,4096,0.008399110701349046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,5120,0.03297777639495002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,3584,0.008271111382378472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,4096,0.029511110650168523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,3584,0.014477334088749357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,3072,0.00756888919406467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,3584,0.0303439994653066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,3072,0.014191110928853353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,2560,0.006275555739800136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,3072,0.029587556918462116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,2560,0.013655111193656921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,2048,0.005324444423119227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,2560,0.02830844455295139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,2048,0.013358222113715278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,1536,0.004827555682924059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,2048,0.027994665834638808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,1536,0.013010666602187686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,1024,0.003969777789380816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,1536,0.026559111144807603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,1024,0.012484444512261285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,768,0.0036195553839206696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,1024,0.025243555506070454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,768,0.012387555506494312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,512,0.003289777785539627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,768,0.02459377712673611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,512,0.011839110818174152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,128,0.01147466649611791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,256,0.0029893333299292457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,512,0.023494222097926672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,256,0.01163822246922387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,64,0.011424889167149862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,128,0.0027804445061418745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,256,0.02288444505797492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,64,0.0025902222841978073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,1024,128,0.022542221678627863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,1024,32,0.00263288886182838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,1024,32,0.011258666714032492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,65536,0.04271555609173245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,65536,0.06169866853290134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,16384,0.014335110783576965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,12288,0.020810667011472914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,16384,0.023287110858493384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,65536,0.11021066374248928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,12288,0.012177777787049612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,16384,0.044682668315039746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,10240,0.011153777440388998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,12288,0.04044622182846069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,10240,0.019647111495335896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,8192,0.010017777482668558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,10240,0.038888000779681735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,8192,0.017264000243610807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,7168,0.009259555902745988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,7168,0.03619911273320516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,5120,0.008594666918118795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,6144,0.01573244399494595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,8192,0.03577955563863119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,7168,0.01752355529202355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,6144,0.008576889004972246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,5120,0.015078221758206686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,4096,0.029318223396937054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,6144,0.03405422303411696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,4096,0.007739555504586961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,5120,0.03130844566557143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,4096,0.01458044515715705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,3584,0.00794933322403166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,3584,0.014418666561444601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,3072,0.007040888898902469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,3584,0.029729776912265356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,3072,0.013983999689420065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,2560,0.006001777946949005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,3072,0.028976887464523315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,2560,0.013550221920013428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,2048,0.0053022222386466126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,2560,0.02833777666091919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,2048,0.013365333278973898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,1536,0.026744888888465032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,1024,0.012636444634861417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,1536,0.004655999855862724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,2048,0.027746667464574177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,1536,0.013042666845851474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,1024,0.003960888832807541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,768,0.0036071112586392295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,768,0.012248888611793518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,1024,0.025219556358125474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,768,0.024917332662476435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,256,0.011765333513418833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,512,0.0033048888047536216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,128,0.011521778172916837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,512,0.01187111106183794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,256,0.0029502221279674103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,512,0.02366488840844896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,128,0.0027644443843099806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,256,0.022711111439598933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,64,0.002624888800912433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,768,128,0.022602667411168415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,64,0.011230222052998014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,768,32,0.0026231110095977783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,768,32,0.011183111204041375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,65536,0.03178222311867608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,65536,0.10966222153769599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,16384,0.023203555080625746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,12288,0.009927999642160203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,65536,0.05824799670113457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,10240,0.009304000271691216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,12288,0.020470221837361652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,16384,0.04430311255984836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,12288,0.0403386652469635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,16384,0.011341333389282227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,8192,0.008653333617581261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,8192,0.03400266501638625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,10240,0.018883556127548218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,8192,0.016711999972661335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,10240,0.03862311111556159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,7168,0.00832177781396442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,5120,0.007380444142553542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,7168,0.016354666815863717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,6144,0.007260444263617198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,6144,0.031962666246626116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,6144,0.015503111812803479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,5120,0.014936889211336771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,4096,0.006404444575309753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,3584,0.013971555564138623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,7168,0.033471110794279314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,5120,0.03053599927160475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,3072,0.013614222407341003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,4096,0.0144195556640625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,3584,0.0070471109615431885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,4096,0.029021332661310833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,3072,0.006545777950021956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,3584,0.029684444268544514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,2560,0.005942222144868638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,2048,0.01311911145846049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,3072,0.028704888290829126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,2560,0.013517333401574029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,2048,0.0052915554907586836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,2560,0.028145776854621038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,1536,0.004620444443490771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,2048,0.027632888820436265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,1024,0.003996444245179494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,1536,0.01276088919904497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,768,0.01218488895230823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,1024,0.012353777885437012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,768,0.003583111282851961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,1536,0.02647022240691715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,1024,0.025081778566042583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,512,0.003321777822242843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,256,0.011599111060301462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,512,0.011858666936556498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,768,0.02446488870514764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,256,0.022732444935374792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,128,0.011495999991893768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,256,0.002943111169669363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,128,0.022672888305452134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,512,512,0.02342133389578925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,128,0.0027715555495686005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,64,0.002622222320901023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,64,0.011167999770906238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,512,32,0.002612444468670421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,65536,0.020800888538360596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,512,32,0.011072888970375061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,65536,0.05663910839292738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,16384,0.008664888640244802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,12288,0.018923555811246235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,16384,0.02199111051029629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,65536,0.11018488804499309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,12288,0.038761778010262385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,12288,0.007858666280905405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,10240,0.03640977872742547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,8192,0.016576889488432143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,16384,0.044008000029457935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,10240,0.0074720001882976955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,10240,0.01759822169939677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,8192,0.007145778172545963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,7168,0.006595555692911148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,7168,0.015828443898095023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,8192,0.03305244445800781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,5120,0.014736889137162102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,6144,0.006542222367392645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,7168,0.033195555210113525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,6144,0.015309333801269531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,5120,0.00675022229552269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,5120,0.030743112166722614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,6144,0.03223022156291538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,4096,0.00628266649113761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,3584,0.0068417779273457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,4096,0.014491554763582019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,3584,0.014053333136770459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,3584,0.02983199887805515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,4096,0.028593778610229492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,3072,0.006447111152940326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,3072,0.013766222529941134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,2560,0.028062221076753404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,2560,0.006010666489601135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,2560,0.013368888861603208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,2048,0.005324444423119227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,3072,0.02900000082121955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,2048,0.013036444783210754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,1536,0.004602666530344221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,2048,0.027450667487250432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,1536,0.012696888711717395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,1024,0.003926222109132343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,512,0.003294222263826264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,1536,0.026143110460705225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,1024,0.01236533373594284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,768,0.003583111282851961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,1024,0.02480711042881012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,768,0.012148444851239523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,768,0.024287111229366724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,512,0.01184977756606208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,256,0.0029226665695508323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,128,0.002744888886809349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,512,0.023345778385798138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,256,0.022815111610624526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,256,0.011605333122942181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,128,0.011535999675591787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,64,0.0025724443710512584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,65536,0.05651199817657471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,16384,0.006596444381607904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,256,128,0.02252977755334642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,16384,0.02123022245036231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,64,0.011239111423492432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,256,32,0.0025964443468385274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,256,32,0.011189333266682096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,65536,0.013704000247849358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,65536,0.10836978091133966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,12288,0.006321777900060018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,16384,0.04182933436499702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,8192,0.0063973334100511335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,12288,0.018792889184421964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,10240,0.006474666711356904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,10240,0.017453332742055256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,12288,0.03774666786193848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,10240,0.036010666026009455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,8192,0.016392888294325936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,8192,0.033406221204333834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,6144,0.01532799998919169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,7168,0.006314666734801398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,7168,0.015862221519152325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,6144,0.006232000059551663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,7168,0.03334488802485996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,5120,0.006538666784763336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,6144,0.03231911195649041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,4096,0.006122666514582104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,3584,0.006576889091067844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,5120,0.014674666855070325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,3072,0.006246222390068903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,4096,0.01440533333354526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,5120,0.030304888884226482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,3584,0.013953778478834363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,4096,0.02892177634769016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,3072,0.013592888911565145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,3584,0.029799110359615747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,2560,0.005949333310127258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,3072,0.02883022361331516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,2560,0.013358222113715278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,2048,0.0052373334765434265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,2560,0.028183112541834515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,1024,0.003926222109132343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,2048,0.012999110751681857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,1536,0.004622222234805425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,2048,0.027053332991070215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,1536,0.012654222548007965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,768,0.003585777762863371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,1024,0.01236533373594284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,1536,0.026013332936498854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,768,0.012133333418104382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,1024,0.024911110599835713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,512,0.0032382222513357797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,768,0.024150222539901733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,512,0.011787555283970304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,256,0.002946666752298673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,512,0.023250665929582384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,256,0.011571555501884885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,128,0.002744888886809349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,64,0.002568888788421949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,128,0.011376000112957425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,128,32,0.0026008888251251644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,64,0.011130666567219628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,256,0.02276977731121911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,32,128,128,0.02239111065864563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,128,32,0.011152889165613385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,12288,0.01866133345497979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,65536,0.012515555653307172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,16384,0.006411555740568373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,65536,0.05486666825082567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,16384,0.021104888783560857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,12288,0.0063031112982167145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,10240,0.006520888871616787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,8192,0.0063680000603199005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,7168,0.01575911045074463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,10240,0.017439999514155917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,7168,0.006250666661394968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,6144,0.006072888771692912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,8192,0.016328889462682936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,5120,0.006412444429265128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,6144,0.015138667490747241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,3584,0.013938667045699226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,5120,0.014592889282438489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,4096,0.014287999934620328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,4096,0.0060159998635451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,3584,0.006477333191368315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,2560,0.013351110948456658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,2048,0.013049778011110095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,3072,0.006249777972698212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,2560,0.005865777946180767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,3072,0.013660444153679742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,2048,0.0052391112678580815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,1536,0.004553777890072929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,1024,0.0039048890272776284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,256,0.002871111035346985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,512,0.011792889071835412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,128,0.002691555561290847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,1536,0.012583110895421771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,768,0.0035777779089079965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,512,0.0032239999208185407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,768,0.012063110868136087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,1024,0.012279999752839407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,256,0.011574221981896294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,64,0.002564444517095884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,128,0.011345778074529437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,64,0.01108266661564509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,64,32,0.002599999929467837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,64,32,0.01108088923825158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,65536,0.011694221860832639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,65536,0.05499733487764994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,16384,0.021213332811991375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,10240,0.017647110753589206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,16384,0.006777777853939269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,10240,0.006583999842405319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,12288,0.006738666858938005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,7168,0.015853333804342482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,12288,0.01869244376818339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,8192,0.006512889017661412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,8192,0.016366221838527255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,5120,0.014651555154058667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,7168,0.006175999840100606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,6144,0.006072888771692912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,5120,0.0064382221963670515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,3072,0.006215111249023014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,6144,0.015216888652907478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,2560,0.005881777654091517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,4096,0.006029333505365584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,3584,0.006414222220579783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,4096,0.014124444789356656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,3584,0.013927110367351107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,3072,0.01365244471364551
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,1024,0.0038924444880750445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,2048,0.005260444349712796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,2560,0.01316088851955202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,2048,0.012983111043771109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,1536,0.004523555437723796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,1536,0.012599111431174807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,768,0.003552000141806073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,1024,0.012264889147546558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,128,0.00272088891102208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,768,0.012098666694429187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,512,0.003216000066863166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,256,0.0028417776856157514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,32,0.002552888873550627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,512,0.011763555308183035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,256,0.011535111400816174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,32,32,64,0.0025742221623659134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,128,0.011391999820868174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,64,0.011175110936164856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,32,32,32,0.011001778145631155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,16384,0.3574693202972412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,12288,0.2742142147488064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,12288,0.5418720245361328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,16384,0.7157635688781738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,10240,0.23062398698594835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,10240,0.45139466391669375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,8192,0.3633626566992865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,8192,0.18858755959404838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,7168,0.1671057807074653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,7168,0.3156382242838542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,6144,0.16625511646270752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,6144,0.3050684399074978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,5120,0.2269351217481825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,5120,0.12509066528744167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,4096,0.18391999933454725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,4096,0.11645421716901992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,8192,0.23778398831685385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,6144,0.18456266985999215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,7168,0.21273333496517607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,10240,0.2857191032833523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,3584,0.09353422456317478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,3584,0.16087555885314941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,3072,0.13808977603912354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,5120,0.14808000458611384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,12288,0.3393830988142226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,3072,0.08294044600592719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,2560,0.1158791118197971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,4096,0.133279111650255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,2560,0.07234844234254625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,2048,0.09274399942821926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,2048,0.06109422445297241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,1536,0.07169422176149157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,1536,0.05032889048258463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,1024,0.048713776800367564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,3584,0.11867822541130914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,1024,0.04040266738997565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,768,0.037843555212020874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,3072,0.10768355263604058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,2560,0.08729155858357747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,2048,0.07605333460701837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,1536,0.06471021970113118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,768,0.036079999473359846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,512,0.02661422226164076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,1024,0.05266133281919691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,768,0.04572622312439812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,512,0.03214577833811442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,256,0.014993778533405729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,128,0.025203555822372437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,256,0.03322844372855292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,64,0.0063635553750726916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,512,0.039523555172814265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,32,0.004620444443490771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,256,0.027292443646325007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,65536,128,0.009712888962692684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,16384,0.40619643529256183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,65536,128,0.029556443293889362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,64,0.022295110755496558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,65536,0.36222222116258407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,65536,32,0.02272533377011617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,16384,0.17379466692606607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,16384,0.10874399873945449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,65536,0.7026533550686307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,12288,0.13285066021813288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,12288,0.09276266892751057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,10240,0.11849066946241592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,10240,0.07922222216924031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,8192,0.0893626676665412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,8192,0.06243377923965454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,7168,0.07852622535493639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,7168,0.055941330062018506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,16384,0.11758666568332249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,12288,0.09776888953314887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,6144,0.07355022430419922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,10240,0.08932444122102524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,6144,0.04860799842410617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,8192,0.07415733072492811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,5120,0.057392888598971896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,5120,0.043042666382259794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,7168,0.07295733027988009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,6144,0.06462933619817098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,4096,0.05448355608516269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,4096,0.037490665912628174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,3584,0.03528177738189697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,3584,0.0413315561082628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,5120,0.05891999933454725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,3072,0.03604177633921305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,4096,0.056530667675866016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,2560,0.030714667505688135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,3072,0.0325564444065094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,3584,0.050154666105906166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,2560,0.02988622254795498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,2048,0.02528800070285797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,1536,0.019985778464211356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,3072,0.046610666645897754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,2048,0.026703110999531213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,2560,0.04387733340263367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,1024,0.014175999495718213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,1536,0.02314222190115187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,2048,0.03990933299064636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,1024,0.03182933396763272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,1024,0.020423110988405015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,768,0.011242667006121742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,1536,0.03547644284036424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,768,0.019296000401178997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,512,0.008477333519193862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,768,0.02994399931695726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,512,0.017672888106769986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,256,0.005790222022268508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,512,0.028384000062942505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,256,0.013564444250530668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,128,0.0034391110142072043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,256,0.024123556084103052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,128,0.013196444345845116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,65536,0.41579198837280273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,64,0.0032097777972618737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,16384,128,0.023437332775857713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,64,0.013213333156373767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,16384,32,0.003144888828198115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,16384,0.13860532972547743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,16384,32,0.013799111048380533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,65536,0.28282401296827525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,16384,0.08254577716191609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,65536,0.5408470895555284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,12288,0.10640533102883233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,12288,0.06597510973612468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,10240,0.08609600199593438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,10240,0.05754933092329237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,8192,0.06969244612587823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,8192,0.04932622114817301
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,7168,0.06111377477645874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,16384,0.10099999772177802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,7168,0.04468889037768046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,12288,0.08611022101508246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,6144,0.053616000546349414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,6144,0.04061244593726264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,8192,0.06988444593217638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,5120,0.044214222166273326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,7168,0.06217510832680596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,5120,0.036481777826944985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,4096,0.032518221272362605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,4096,0.036101334624820285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,6144,0.056840889983707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,5120,0.05240355597601997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,3584,0.0326551099618276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,3584,0.030439999368455675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,4096,0.04766311248143514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,3072,0.02812533246146308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,10240,0.19578044944339326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,3584,0.04518933428658379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,3072,0.028044445647133723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,2560,0.026173333326975506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,3072,0.04257422354486254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,2560,0.026100445124838088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,2048,0.03548533386654324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,2560,0.039934221241209246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,2048,0.023287110858493384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,1536,0.016707556115256417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,1024,0.019273777802785236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,2048,0.0361697773138682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,1536,0.021311110920376245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,1024,0.01874755488501655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,65536,0.3442408773634169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,1536,0.03299644589424133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,768,0.009920000202125972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,768,0.018168000711335074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,1024,0.030343112018373277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,512,0.009168888959619734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,768,0.02880177895228068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,512,0.015408888459205627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,256,0.0041680000722408295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,512,0.027615111735132005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,256,0.013252444565296173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,128,0.023045332895384893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,128,0.004516444272465176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,12288,256,0.023751111494170293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,128,0.01291911138428582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,64,0.0031226666437255014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,64,0.012979555461141797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,12288,32,0.0031840000301599503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,12288,32,0.013592888911565145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,65536,0.24853600396050346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,16384,0.11288532945844863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,16384,0.07298133108350965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,65536,0.5007040235731337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,12288,0.08639199866188897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,12288,0.05818844503826565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,10240,0.07245510816574097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,10240,0.0897795557975769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,8192,0.05843466520309448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,8192,0.04407555527157254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,16384,0.09003021981981064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,7168,0.05157155460781521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,12288,0.07657955752478705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,7168,0.042471110820770264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,6144,0.045533332559797496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,10240,0.06769600179460314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,6144,0.0370399985048506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,5120,0.03804355528619554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,5120,0.03378399875428941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,4096,0.031035555733574763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,6144,0.052777777115503945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,7168,0.056623107857174344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,4096,0.02994222111172146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,3584,0.027652444110976324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,5120,0.04864355590608385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,3584,0.028229332632488672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,4096,0.044863111442989774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,3072,0.024199111594094172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,3072,0.026260443859630164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,3584,0.04296888907750448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,8192,0.15808355808258057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,2560,0.020755555894639757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,2560,0.02420444455411699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,3072,0.039890666802724205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,2048,0.017072000437312655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,1536,0.020459555917316012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,2048,0.022243556049134996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,2560,0.037363555696275495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,1536,0.013384000294738345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,65536,0.3110995557573107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,2048,0.06568977567884657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,1024,0.009911110831631554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,1536,0.03241511185963949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,1024,0.018592000007629395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,768,0.008280889027648503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,1024,0.029770665698581274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,768,0.01755644381046295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,256,0.013016888664828407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,512,0.008167111211352879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,768,0.028951111767027114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,512,0.015072888798183866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,256,0.0035822221802340615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,512,0.02696977721320258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,128,0.0031511110977994073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,256,0.023520888553725347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,32,0.013418667018413544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,128,0.012809777425395118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,64,0.002943111169669363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,10240,128,0.02979555394914415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,10240,64,0.01296266665061315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,10240,32,0.002973333415057924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,65536,0.1932755576239692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,12288,0.06793689065509372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,65536,0.35674312379625106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,16384,0.08847733338673909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,10240,0.05690755446751913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,16384,0.059752000702752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,12288,0.049117333359188504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,10240,0.043064001533720225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,8192,0.03761511047681173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,8192,0.04628888765970866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,12288,0.07168000274234347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,7168,0.04097244474622939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,10240,0.0606364475356208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,7168,0.034471998612085976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,6144,0.0362426671716902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,8192,0.052899556027518384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,6144,0.03444088829888238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,7168,0.05166133244832357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,5120,0.030566222137875024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,5120,0.028908444775475398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,6144,0.048067556487189404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,4096,0.02507200009293026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,16384,0.16812888781229654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,5120,0.04460444384151035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,4096,0.025292444560262892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,3584,0.022336888644430373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,3584,0.023727999793158636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,3072,0.021693333983421326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,4096,0.04152444336149428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,65536,0.2565760082668728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,3072,0.01977422171168857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,3072,0.03707733419206407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,3584,0.058409776952531606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,2048,0.016004444824324716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,2560,0.01682577696111467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,2560,0.02030044462945726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,2560,0.03486844566133287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,2048,0.019009777241282992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,1536,0.014368888404634265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,1024,0.015726221932305228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,2048,0.03299199872546726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,1536,0.017136888371573556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,1024,0.00847999999920527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,1536,0.030949334303538006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,768,0.008124444219801161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,768,0.013762666947311826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,1024,0.02875111169285244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,512,0.0057591112951437635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,768,0.027581334114074707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,512,0.012186666329701742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,256,0.02350666622320811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,256,0.0037537775933742523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,256,0.011848889291286469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,512,0.0247999992635515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,128,0.003084444544381566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,32,0.012186666329701742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,128,0.011505777637163797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,64,0.002934222213096089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,8192,128,0.02307644486427307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,16384,0.08223733637068006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,8192,64,0.011591110792424945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,8192,32,0.003003555453485913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,12288,0.06406666835149129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,65536,0.17502755588955352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,65536,0.30665066507127553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,16384,0.05492711067199707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,12288,0.045447998576694064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,10240,0.0543759995036655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,16384,0.07443288962046306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,12288,0.06350133154127333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,8192,0.0351893338892195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,10240,0.04174222217665779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,8192,0.04143111242188348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,7168,0.03677777780426873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,10240,0.05788089169396294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,7168,0.03227377931276957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,6144,0.03233155608177185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,8192,0.053632001082102455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,6144,0.029674665795432195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,7168,0.049011554982927114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,5120,0.02755022214518653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,5120,0.027237332529491846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,4096,0.02351644469632043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,6144,0.04695555567741394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,4096,0.022721777359644573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,5120,0.043635557095209755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,3584,0.02038666605949402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,4096,0.03968888852331374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,3584,0.02232711182700263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,65536,0.23156889279683432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,2560,0.019671110643280875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,3072,0.018073777357737224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,3584,0.03781066669358148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,3072,0.02054755555258857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,1536,0.010469333165221743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,2560,0.015455999308162265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,3072,0.03527911172972785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,2048,0.012812444733248817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,2560,0.03428444597456191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,1024,0.015494222442309061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,2048,0.018198221921920776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,2048,0.03237333231502109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,1536,0.016725333200560678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,1024,0.007912889122962952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,1536,0.030674666166305542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,768,0.006768888897365994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,1024,0.02910311023394267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,768,0.012515555653307172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,512,0.005335111171007156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,768,0.02737244466940562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,512,0.012078222301271228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,256,0.0033288889875014624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,64,0.0028391112056043413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,512,0.02402755618095398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,256,0.011734222372372946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,128,0.003036444385846456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,256,0.023079999619060095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,128,0.011532444092962476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,7168,128,0.023072888453801472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,7168,32,0.002847111059559716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,64,0.011560888753996955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,7168,32,0.012087110843923358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,65536,0.1601573361290826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,65536,0.26762400733100045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,12288,0.05543377664354113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,16384,0.07334222396214803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,16384,0.05080622103479174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,12288,0.041879110866122775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,10240,0.0458586679564582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,10240,0.03751911057366265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,16384,0.07055199808544584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,8192,0.03807199994723002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,12288,0.06272711356480916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,8192,0.03311466508441501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,10240,0.054839111036724515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,7168,0.03349422083960639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,7168,0.030555556217829388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,8192,0.04849511053827074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,6144,0.029390222496456567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,6144,0.028139554791980322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,7168,0.05304977628919813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,5120,0.024873778223991394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,6144,0.04513066675927904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,5120,0.02553066611289978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,65536,0.2037688891092936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,4096,0.0210524449745814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,4096,0.037950221035215594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,4096,0.02382044494152069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,5120,0.04141689009136624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,3584,0.0188426673412323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,3584,0.02113866640461816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,3072,0.016536888149049547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,3584,0.03686577743954129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,3072,0.019959110352728102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,2560,0.016034666034910414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,3072,0.03541333476702372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,2560,0.019163555569118924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,2048,0.012230222423871359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,2560,0.033222221665912204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,2048,0.01756177842617035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,1536,0.00979555563794242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,2048,0.03187200095918443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,1536,0.016348444753222995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,1024,0.007394666473070781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,512,0.0041999999019834726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,1536,0.030584888325797185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,1024,0.013818666338920593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,768,0.006234666539563074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,1024,0.027983999914593164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,768,0.012311999996503195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,768,0.02534577747186025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,512,0.012273777690198688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,128,0.011420444481902652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,256,0.0032791110376516977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,512,0.023975999818907842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,256,0.011785777906576792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,128,0.003026666740576426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,32,0.012014221813943652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,256,0.02336799932850732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,64,0.0028728888266616394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,6144,128,0.022978666755888198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,6144,64,0.011395555403497485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,6144,32,0.0028408887899584244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,65536,0.13774222797817653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,12288,0.044714666075176664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,65536,0.2232044537862142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,16384,0.05818577607472738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,16384,0.045490665568245776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,16384,0.06767821974224515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,10240,0.03393866618474325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,12288,0.03771911064783732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,10240,0.037951111793518066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,8192,0.03085955646302965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,12288,0.05717155668470594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,8192,0.030202666918436687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,10240,0.05376266770892673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,7168,0.027449778384632532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,65536,0.18090221616956922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,7168,0.027975112199783325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,6144,0.02440266642305586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,8192,0.049359109666612416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,6144,0.02562933332390255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,7168,0.045977777904934354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,5120,0.020747555626763236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,3584,0.015615110596021017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,6144,0.0432640016078949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,5120,0.024708444873491924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,4096,0.017343110508388944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,5120,0.04014399978849623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,4096,0.020853334003024634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,4096,0.03808355662557814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,3584,0.019984000258975558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,3072,0.03418044580353631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,3072,0.013602666556835175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,3584,0.03606311149067349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,3072,0.018764444523387484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,2560,0.011887999872366587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,2560,0.017920888132519193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,2048,0.010097777677906884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,2048,0.017024889588356018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,2560,0.033201777272754245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,1536,0.008476444416575963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,2048,0.03123733401298523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,1536,0.015801777442296345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,1024,0.006589333216349284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,1536,0.029987557066811457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,1024,0.012679110798570844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,768,0.005671111245950063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,1024,0.027290667096773785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,768,0.012303111453851065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,512,0.003650666524966558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,256,0.011722666521867117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,512,0.011961778004964193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,768,0.024562666813532513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,256,0.003185777821474605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,512,0.023967999551031325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,128,0.002921777880854077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,128,0.011370666325092316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,256,0.02327288852797614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,32,0.01198755535814497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,64,0.0027599999060233435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,5120,128,0.022760889596409265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,5120,64,0.011405333048767514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,5120,32,0.002798222212327851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,65536,0.17745510737101236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,16384,0.04906577865282694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,12288,0.03873066769705878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,65536,0.11422933472527398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,16384,0.04000088903639052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,12288,0.03431111243036058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,16384,0.06334488921695285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,10240,0.03309422069125705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,10240,0.029926220575968426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,8192,0.027639999985694885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,12288,0.054586665497885815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,8192,0.027241778042581346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,7168,0.024951110283533733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,65536,0.17054488923814562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,10240,0.05159733361668057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,8192,0.04449511236614651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,7168,0.025076443950335186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,6144,0.022250667214393616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,6144,0.022794667217466567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,7168,0.04704711172315809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,4096,0.019520888725916546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,5120,0.017628444565667045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,6144,0.04148533278041416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,3584,0.018664000762833487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,5120,0.020971554848882887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,4096,0.01629688839117686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,5120,0.03861155443721347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,3584,0.015056000815497505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,4096,0.036566221051745944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,3072,0.011624889241324531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,3072,0.017765333255132038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,2560,0.03267200125588311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,3584,0.03482044405407376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,2560,0.01016266644001007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,3072,0.03324088785383437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,2560,0.017183999220530193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,1024,0.006019555446174409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,2048,0.008763555851247575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,2048,0.016196444630622864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,1536,0.007308444215191736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,2048,0.030807110998365615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,1536,0.014228444960382251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,1024,0.01257599973016315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,1536,0.029027554723951552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,768,0.004563555535342959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,1024,0.025462221768167283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,768,0.012297777665985955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,512,0.0034577778230110803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,768,0.02458933326933119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,512,0.011965333587593503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,128,0.011324444578753578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,256,0.003089777711364958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,512,0.024088889360427856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,32,0.0027520000520679685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,256,0.011674666570292579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,128,0.0028986665937635633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,256,0.023099554909600153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,4096,64,0.002736888825893402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,4096,128,0.023028444912698533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,64,0.011343110766675739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,4096,32,0.011976000335481433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,65536,0.1626124382019043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,65536,0.10212000211079915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,16384,0.04399199949370491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,12288,0.034308443466822304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,16384,0.03672088848219977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,12288,0.031013333135181006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,10240,0.0298942228158315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,10240,0.028113779094484117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,8192,0.025048888391918604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,10240,0.04948177933692932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,12288,0.0534746679994795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,16384,0.06298044655058119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,8192,0.024853333830833435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,65536,0.15799821747673884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,7168,0.02276888820860121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,7168,0.022987556126382615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,6144,0.022641777992248535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,8192,0.04323733184072706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,6144,0.022054221895005967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,7168,0.04289155536227756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,6144,0.040355556541019015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,5120,0.0380657778845893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,5120,0.0176862229903539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,5120,0.020181333025296528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,4096,0.014905777242448596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,4096,0.019153777096006606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,3584,0.013619555367363824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,4096,0.0358568893538581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,3584,0.017811555001470778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,3072,0.010688888529936472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,3584,0.03467377689149644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,3072,0.01757333344883389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,2560,0.009363555245929295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,3072,0.0335440006521013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,2560,0.016780444317393832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,2048,0.008220444122950235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,2560,0.032199111249711775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,2048,0.01568000018596649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,1536,0.0070426662762959795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,2048,0.030693332354227703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,768,0.00378222225440873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,1536,0.013907555076811047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,1024,0.0058613332609335584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,1536,0.027672888504134283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,1024,0.012411555482281579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,1024,0.025389333566029865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,768,0.012191111014948951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,512,0.0037671112351947357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,768,0.024196444286240473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,512,0.012072888513406118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,256,0.003072888900836309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,512,0.023784889115227595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,256,0.011519999967681037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,128,0.002842666581273079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,256,0.022950222094853718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,128,0.01132266637351778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,64,0.002741333304180039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3584,128,0.022876444790098403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,64,0.011296889020336999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,16384,0.0392871101697286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3584,32,0.002798222212327851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3584,32,0.011788444386588203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,65536,0.13876356018914118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,65536,0.089355554845598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,16384,0.03350399931271871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,12288,0.030919111437267725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,12288,0.028432889117134943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,16384,0.06033955679999458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,10240,0.027224888404210407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,10240,0.0255724440018336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,7168,0.021074665917290583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,12288,0.052417778306537204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,8192,0.023374223046832617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,8192,0.02294933299223582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,10240,0.04870577653249105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,65536,0.1800586647457547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,8192,0.04320799973275927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,7168,0.02134399943881565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,6144,0.04010844561788771
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,5120,0.01979022224744161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,6144,0.018806222412321303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,6144,0.02054666644997067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,7168,0.042261332273483276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,5120,0.016303999556435477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,4096,0.014339556296666464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,4096,0.018145778112941317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,3584,0.03487822082307603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,5120,0.037864890363481306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,3584,0.013083555632167392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,4096,0.03632888860172696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,3584,0.017792888813548617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,3072,0.011946666571829053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,3072,0.016943999462657504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,2560,0.008810666700204214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,3072,0.0328746669822269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,2560,0.016399999459584553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,2048,0.007506666911972894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,2560,0.03189244535234239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,2048,0.014314666390419006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,1024,0.012490666574902005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,1536,0.00666222224632899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,768,0.0037804444630940757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,768,0.012248000337017907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,2048,0.03048444456524319
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,1536,0.013057777451144325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,768,0.02425777746571435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,512,0.012033777932325998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,1024,0.005046222358942032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,1536,0.026700443691677515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,1024,0.024939555260870192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,512,0.0034666665726237824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,256,0.003104000041882197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,512,0.02365600069363912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,256,0.011511111425028907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,128,0.02278311053911845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,128,0.0028693332440323303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,3072,256,0.023139556248982746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,128,0.011396444506115384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,64,0.0027591110103660156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,64,0.01128000020980835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,3072,32,0.0027386666172080566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,3072,32,0.011772444678677453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,16384,0.030616889397303265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,65536,0.11629688739776611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,65536,0.0794888867272271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,16384,0.03330311179161072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,12288,0.026724444495307073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,12288,0.026191999514897663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,10240,0.02327377763059404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,16384,0.05902577771080864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,10240,0.023353776997990076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,12288,0.051445331838395864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,8192,0.019532443748580087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,65536,0.1361128886540731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,8192,0.02115733259254032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,10240,0.04750400119357639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,7168,0.017812444104088675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,8192,0.04169244567553202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,7168,0.020278222031063504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,6144,0.015821332732836407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,7168,0.04110399881998698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,6144,0.019380443625979953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,5120,0.014173333843549093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,6144,0.03990933299064636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,5120,0.018259555101394653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,4096,0.012179555164443122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,3584,0.017027555240525138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,4096,0.017768889665603638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,5120,0.03770133190684848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,3584,0.011485333244005838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,3072,0.010620444185203977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,4096,0.035946667194366455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,3072,0.01621866722901662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,3584,0.034373333056767784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,2560,0.008220444122950235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,3072,0.03351644343800015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,2560,0.015108444624476962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,2048,0.007094221810499827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,2560,0.03193333413865831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,1536,0.026965333355797663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,2048,0.013234666652149625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,1536,0.006055111272467508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,2048,0.02812355425622728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,1536,0.0129457778400845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,1024,0.004531555705600315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,1024,0.012409778104888068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,768,0.003738666574160258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,1024,0.025016888976097107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,768,0.012072888513406118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,512,0.0034168888297345904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,768,0.024099555280473497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,512,0.011902222202883827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,256,0.0030622221529483795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,512,0.023418666587935552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,256,0.011520889070298938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,128,0.002830222249031067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,256,0.02307111190425025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,128,0.011368888947698804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,64,0.0027057776848475137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2560,128,0.022488888767030504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,64,0.011236444115638733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2560,32,0.0027351110345787476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,16384,0.028040889236662123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2560,32,0.011736888852384357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,65536,0.09276088741090562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,65536,0.07096799876954821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,16384,0.02742044462098016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,12288,0.02225688927703434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,12288,0.023387556274731953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,16384,0.05598666932847765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,10240,0.01940444442960951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,8192,0.020000000794728596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,12288,0.04962844318813748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,65536,0.1440568897459242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,10240,0.021922666165563796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,8192,0.016137777103318107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,10240,0.045924444993337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,6144,0.018631110588709515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,7168,0.014856000741322836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,8192,0.040968888335757785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,7168,0.01923466722170512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,4096,0.010632000035709806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,6144,0.013336000343163809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,7168,0.0402693317996131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,5120,0.012285333540704517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,6144,0.038336889611350164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,5120,0.017714665995703805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,4096,0.016738666428460013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,5120,0.037045333120558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,3584,0.010120888551076254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,4096,0.03483555714289347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,3584,0.015568888849682279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,3072,0.009334222310119206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,3584,0.033844444486829973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,3072,0.014907555447684394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,2560,0.007368000017272101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,3072,0.03228444523281521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,2560,0.014090667168299357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,2048,0.006586666736337874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,2560,0.0309840010272132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,1024,0.003955555458863576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,2048,0.013231111069520315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,1024,0.012451555165979596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,1536,0.005513777749405966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,2048,0.027810666296217177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,1536,0.012703110774358114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,1536,0.026354667213228013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,768,0.0038604442444112566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,1024,0.0247022228108512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,768,0.012162666353914471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,512,0.0033644443998734155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,768,0.024075556132528517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,512,0.011785777906576792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,256,0.002990222225586573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,512,0.0236435549126731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,256,0.0114702218108707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,128,0.0028008888992998335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,256,0.023031999667485554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,128,0.011189333266682096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,32,0.011783999701340994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,64,0.002611555573013094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,2048,64,0.011329777538776398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,2048,32,0.0026088888860411118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,2048,128,0.022616000639067754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,65536,0.07125955820083618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,65536,0.06544799937142266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,16384,0.022447110878096685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,16384,0.026169778572188482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,12288,0.018201778332392376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,12288,0.04760711060629951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,12288,0.02219377789232466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,10240,0.016154666741689045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,16384,0.054357333315743335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,65536,0.1344266732533773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,10240,0.02090222140153249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,8192,0.01368533323208491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,10240,0.044975999328825206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,8192,0.019371555911170114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,7168,0.012741333080662621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,8192,0.04165777895185683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,7168,0.01850400037235684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,5120,0.017156443662113614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,6144,0.011776888536082374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,7168,0.040022220876481794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,6144,0.017821333474583097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,3584,0.009069333473841349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,3584,0.014332445131407844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,5120,0.010760000182522668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,6144,0.03806133402718438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,4096,0.009655111365848118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,5120,0.03642222285270691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,4096,0.015281778242852954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,4096,0.03457066747877333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,3072,0.008662222160233391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,3584,0.03343111276626587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,3072,0.01385155568520228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,2560,0.006934222247865465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,3072,0.030477331744299993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,2560,0.013381332986884646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,2048,0.006308444258239534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,2560,0.02897600001758999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,1536,0.026296888788541157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,2048,0.012991999586423239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,1536,0.004651555584536659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,2048,0.027735999888843958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,768,0.012055111428101858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,1536,0.012679999901188744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,1024,0.003997333347797394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,1024,0.012342222034931183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,768,0.003683555457327101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,1024,0.02456444501876831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,512,0.0033244445092148254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,768,0.02390488816632165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,512,0.011858666936556498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,256,0.002966222249799304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,512,0.023611555496851604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,256,0.011557333171367645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,128,0.002777777819169892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,256,0.022856000396940444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,32,0.011678222152921887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,128,0.011377777490350934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,64,0.0026319999661710528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1536,64,0.011117333339320289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1536,32,0.00259911103381051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1536,128,0.02273688879277971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,65536,0.049104889233907066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,65536,0.055567112233903676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,65536,0.11533510684967041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,16384,0.01658933361371358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,16384,0.023218666513760883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,12288,0.014097778333557976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,12288,0.020502222908867728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,16384,0.04826755656136406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,10240,0.012562666502263812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,10240,0.01961866683430142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,12288,0.045111109813054405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,8192,0.011057777537239922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,7168,0.0178195552693473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,10240,0.04219911164707608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,8192,0.018366222580273945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,6144,0.016629333297411602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,7168,0.010300444232092964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,8192,0.03797422183884515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,6144,0.009331555830107795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,7168,0.03815733393033346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,5120,0.008986666798591614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,5120,0.01498933302031623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,6144,0.03640177845954895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,4096,0.00832622249921163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,4096,0.014511111709806653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,5120,0.034895999564064875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,3584,0.00814488861295912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,4096,0.0325271122985416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,3584,0.014179555906189812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,3072,0.007382222347789341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,3584,0.03109688891304864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,3072,0.013930666777822705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,2560,0.006331555545330048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,3072,0.029263112280103896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,2560,0.013471110827393003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,2048,0.005600888695981767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,1536,0.012663111090660095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,2560,0.027758222487237718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,2048,0.013200889031092325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,1536,0.00463733325401942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,2048,0.026895999908447266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,1024,0.00396799999806616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,1024,0.01223288890388277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,1536,0.0255039996571011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,768,0.003653333418899112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,1024,0.02425333360830943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,768,0.012223111258612739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,512,0.0232640008131663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,512,0.00332266671790017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,768,0.023749333288934495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,512,0.011920000116030375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,256,0.002962666667169995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,256,0.011623111036088733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,128,0.0027955555253558685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,256,0.0229111115137736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,32,0.011703111231327057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,128,0.01128888875246048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,64,0.0026168889469570587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,1024,128,0.02254399988386366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,1024,64,0.011289777855078379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,1024,32,0.002630222174856398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,65536,0.03897777862019009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,65536,0.107113778591156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,65536,0.05195733242564731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,16384,0.014363555444611443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,16384,0.021919111410776775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,12288,0.012014221813943652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,16384,0.04453866680463155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,12288,0.019867555962668527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,8192,0.01715377800994449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,10240,0.011029333704047732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,12288,0.04250133368704054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,10240,0.01897333396805657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,8192,0.010091555615266165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,7168,0.03610488772392273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,10240,0.03934577769703335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,7168,0.009308444129096137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,8192,0.03544444508022732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,7168,0.01629955569903056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,6144,0.008354666332403818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,6144,0.015370666980743408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,4096,0.014373333917723762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,5120,0.008583111067612966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,6144,0.03432622220781114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,4096,0.02944799926545885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,5120,0.014952000644471912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,4096,0.007601777712504069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,5120,0.031410667631361223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,3584,0.007527111305130853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,3584,0.01405066748460134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,3072,0.0068604445291890045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,3072,0.013905778527259827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,3584,0.029064890411165025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,2560,0.00601333338353369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,3072,0.02769600020514594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,2560,0.013427555561065674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,2048,0.02625422179698944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,2048,0.0053022222386466126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,2560,0.0270506673389011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,2048,0.013005333642164866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,1536,0.004690666579537922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,1536,0.012678222523795234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,1024,0.003953777667548922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,1536,0.02533600065443251
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,1024,0.012260444462299347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,768,0.0036186666952239144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,1024,0.02397777802414364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,768,0.011977777712874942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,512,0.0032746667663256326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,768,0.023787554767396715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,512,0.011804444094498953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,256,0.002983999955985281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,128,0.011400000088744693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,512,0.023208889696333144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,64,0.002647111192345619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,256,0.011589333415031433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,32,0.0026106666773557663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,768,128,0.0027431110954946945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,256,0.02276355524857839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,768,128,0.022487999664412603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,64,0.011197333534558615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,768,32,0.011735111474990845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,65536,0.028726223442289565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,65536,0.04968800147374471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,16384,0.011388444238238864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,16384,0.021687111920780603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,65536,0.09822844134436713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,12288,0.009481777747472128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,12288,0.039954665634367205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,16384,0.043640888399548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,12288,0.019575999842749704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,10240,0.009568000005351173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,10240,0.018171555466122098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,8192,0.008435555630260045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,7168,0.015772443678643968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,10240,0.03673600157101949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,8192,0.01662133302953508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,7168,0.008067555725574493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,8192,0.033082667324278094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,6144,0.007724444071451823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,7168,0.03311999970012241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,6144,0.015214223000738354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,5120,0.007128000259399414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,6144,0.03199822372860379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,5120,0.014911999305089315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,4096,0.006627555522653792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,5120,0.030592888593673706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,4096,0.014379555980364481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,3072,0.013938667045699226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,3584,0.0069475554757648045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,3072,0.027831110689375136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,4096,0.029534220695495605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,3584,0.014064888159434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,3584,0.028920888900756836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,2048,0.013086222112178802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,3072,0.006548444430033366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,2560,0.005977777971161737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,1536,0.012697777814335294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,2560,0.013637333280510373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,2048,0.0053173332578606075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,2560,0.027116444375779893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,1536,0.004619555754794015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,2048,0.026120889517996047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,1024,0.003960888832807541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,1024,0.01221866657336553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,1536,0.025449777642885845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,768,0.003608889049953885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,1024,0.024179556303554114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,768,0.01204177737236023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,512,0.003305777700410949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,768,0.023573334018389385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,512,0.011855111353927188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,256,0.0029519999192820657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,512,0.023009777069091797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,256,0.011585777832402123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,64,0.011137777732478248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,128,0.0027511111564106415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,256,0.022851554883850947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,128,0.011281777587201862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,65536,0.048431111706627734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,64,0.002628444383541743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,512,128,0.022574222750133936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,65536,0.09772177537282307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,512,32,0.00258933338854048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,512,32,0.01167555567291048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,65536,0.017549332645204332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,16384,0.008347555167145198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,16384,0.020974222156736586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,10240,0.01720177796151903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,12288,0.007678222325113084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,10240,0.03544977638456557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,16384,0.04253777861595154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,12288,0.018415111634466384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,10240,0.007063999772071838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,12288,0.038338667816585965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,8192,0.00684799998998642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,8192,0.016302223006884258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,7168,0.0063999998900625445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,5120,0.006800888727108638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,8192,0.03271022107866075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,6144,0.03160266743765937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,7168,0.01585244470172458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,6144,0.006487111250559489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,4096,0.014289778139856128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,7168,0.03283910950024923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,6144,0.015230221880806817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,5120,0.014828445182906257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,3072,0.0063991112013657885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,4096,0.00628266649113761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,5120,0.030371556679407757
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,3584,0.006775110960006714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,2560,0.005952888892756567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,3072,0.02769244379467434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,2560,0.013415111435784234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,4096,0.029083556599087183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,3584,0.013964444398880005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,3584,0.028773334291246202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,3072,0.013784888717863293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,2048,0.0052373334765434265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,1536,0.012778666284349231
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,2560,0.02683466672897339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,2048,0.013125333521101209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,1024,0.012406222522258759
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,1536,0.02514044443766276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,1536,0.004587555511130227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,768,0.012190221912331052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,1024,0.023967110448413428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,2048,0.025923555095990498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,512,0.011970666547616323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,1024,0.003936888857020272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,768,0.0035422220826148987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,512,0.0032684444967243406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,768,0.023400889502631292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,256,0.0029395555870400537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,512,0.0229093333085378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,256,0.01162577751610014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,128,0.0027191111197074256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,256,0.022521777285469904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,128,0.011205332974592844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,64,0.002605333303411802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,256,128,0.022311111291249592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,64,0.011158222125636207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,256,32,0.002604444407754474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,256,32,0.01166311071978675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,65536,0.012505778008037143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,65536,0.047672000196244985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,16384,0.006561777657932705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,16384,0.020061333974202473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,12288,0.006633777585294511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,65536,0.09637511438793606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,16384,0.04045155644416809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,12288,0.018106665876176622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,8192,0.016227554943826463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,10240,0.006897777732875612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,12288,0.037853333685133196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,7168,0.015711999601787992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,10240,0.016935999194780987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,8192,0.00656355544924736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,10240,0.035190221336152815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,7168,0.0063591111037466265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,8192,0.032615999380747475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,6144,0.00608800020482805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,6144,0.015207111835479736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,7168,0.032609777318106756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,5120,0.006506666541099548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,6144,0.03169066707293192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,5120,0.014650666051440768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,4096,0.006075555665625467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,5120,0.03046222196684943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,4096,0.014208000567224292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,3584,0.006477333191368315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,4096,0.02906755606333415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,3584,0.013912000589900546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,3072,0.006147555592987273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,3584,0.028780443800820246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,2560,0.026644445127911035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,3072,0.01349422252840466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,2560,0.005683555371231503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,3072,0.027424888478385076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,2560,0.013120888835854001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,2048,0.005104889058404499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,2048,0.012901333471139273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,1536,0.004488000025351842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,1536,0.012596444123321109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,2048,0.025949334104855854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,1024,0.003847111016511917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,1536,0.024829333027203877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,1024,0.012087999946541257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,768,0.003519111209445529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,1024,0.023912888434198167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,768,0.012039999994966718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,512,0.003175111073586676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,768,0.023437332775857713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,512,0.011705777711338468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,256,0.0028773333049482773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,512,0.022840888963805303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,256,0.01143022212717268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,128,0.0027342221389214196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,256,0.02277155551645491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,128,0.011214222345087262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,64,0.0025840000145965153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,16,128,128,0.02241244415442149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,128,32,0.00254844439526399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,64,0.011152000062995486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,128,32,0.01167555567291048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,65536,0.00979466653532452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,65536,0.04735733403099907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,16384,0.006380444599522485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,16384,0.019703111714786954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,12288,0.006268444574541516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,12288,0.01756533318095737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,10240,0.0063279999627007385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,10240,0.01680088871055179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,8192,0.006228444476922353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,8192,0.0159244438012441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,7168,0.006093333164850871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,7168,0.015407999356587728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,6144,0.005964444329341252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,6144,0.014741332994567024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,5120,0.006330666856633292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,3072,0.00610133343272739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,5120,0.0143857780430052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,4096,0.006019555446174409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,3584,0.0064239998658498125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,4096,0.013883555928866068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,3584,0.013596444494194455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,2048,0.012688889271683164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,1536,0.004376000000370873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,2560,0.005585777676767773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,3072,0.01330933305952284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,2560,0.013003555436929068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,2048,0.005021333280536863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,1024,0.0037360000941488477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,1536,0.012341332932313284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,1024,0.011972444752852121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,768,0.0034302222645945023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,512,0.0031617778456873367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,128,0.011136000355084738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,768,0.011890666352377998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,64,0.011194667054547204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,512,0.011717333561844297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,32,0.011635555161370171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,256,0.0028320000403457214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,64,256,0.011417778001891242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,128,0.002620444529586368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,64,0.002532444480392668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,64,32,0.002522666628162066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,65536,0.009223110973834991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,65536,0.04630400074852837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,16384,0.006486222147941589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,16384,0.019792889555295307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,7168,0.006063999815119638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,12288,0.0063688887490166565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,12288,0.017703110973040264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,10240,0.006266666783226862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,10240,0.016789333687888253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,8192,0.006177777631415262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,8192,0.01610133383009169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,7168,0.015447111593352424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,6144,0.00591822216908137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,6144,0.015056888262430826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,3072,0.005995555470387141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,5120,0.006267555471923616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,5120,0.014325333966149224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,4096,0.0058959999846087555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,2560,0.013056000073750814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,3584,0.0064195555945237475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,4096,0.014019555515713163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,3584,0.013667555318938361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,2560,0.005572444448868434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,3072,0.01332355539004008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,2048,0.004970666848950916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,2048,0.012525333298577202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,512,0.0031662223239739737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,1536,0.004373333520359463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,512,0.011595555477672152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,1536,0.012336889074908363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,256,0.011333333121405708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,1024,0.0037466668420367767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,64,0.002529777793420686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,1024,0.01203644441233741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,768,0.003407111184464561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,768,0.011868444581826528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,256,0.0028311111446883944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,128,0.002663111107216941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,128,0.011125333607196808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,64,0.011137777732478248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,16,32,32,0.002528000002106031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,16,32,32,0.011607111328177981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,16384,0.353965335422092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,12288,0.2712693214416504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,16384,0.7151288986206055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,12288,0.540795538160536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,10240,0.2281573348575168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,10240,0.45063469145033097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,8192,0.21443733904096815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,8192,0.36424711015489364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,7168,0.16466932826571995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,7168,0.3152959876590305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,6144,0.271126217312283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,6144,0.15438844097985163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,5120,0.2258080111609565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,5120,0.12310666508144802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,4096,0.18206844064924452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,4096,0.11243200302124023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,6144,0.18285511599646675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,8192,0.23455466164482963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,7168,0.19665955172644722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,10240,0.28803467750549316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,3584,0.1594826645321316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,5120,0.15869866477118597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,3584,0.09189955393473308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,12288,0.33850577142503524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,3072,0.15788888931274414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,3072,0.08137777778837416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,4096,0.13478844695621067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,2560,0.12661777602301702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,2560,0.0709457794825236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,2048,0.059876441955566406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,2048,0.10833599832322861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,3584,0.1060604453086853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,16384,0.400548431608412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,1536,0.07065244515736897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,1536,0.05021244287490845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,3072,0.10577066739400227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,1024,0.05113333463668823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,2560,0.09063111411200629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,1024,0.04127111037572225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,768,0.03642488850487603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,2048,0.07290222247441609
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,768,0.03761511047681173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,1536,0.06722488668229845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,512,0.0258933338854048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,1024,0.05531555414199829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,512,0.03196444445186191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,768,0.045829332537121244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,256,0.014599111345079211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,64,0.006285333385070165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,512,0.03887466589609782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,128,0.008789333204428354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,256,0.028260442945692275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,32,0.021835555632909138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,256,0.03281955586539374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,128,0.02550044490231408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,65536,32,0.004460444466935264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,65536,128,0.02985244327121311
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,65536,64,0.022416000564893086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,16384,0.1744497749540541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,65536,0.3584871027204726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,16384,0.10060177909003364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,12288,0.1332133346133762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,65536,0.7003342310587565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,12288,0.0804319977760315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,10240,0.11012622382905747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,10240,0.06926666365729438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,8192,0.08916622400283813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,8192,0.058393776416778564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,7168,0.07840622133678861
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,7168,0.05307111144065857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,16384,0.12850399812062582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,12288,0.10827377769682144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,6144,0.05310666561126709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,6144,0.06873599688212077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,10240,0.08571733368767633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,8192,0.08179733488294813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,5120,0.05722044573889839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,7168,0.06914755370881823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,5120,0.04239644606908163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,4096,0.04650755392180549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,4096,0.03684355484114753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,6144,0.06340266598595513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,3584,0.041174223025639854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,5120,0.0676293306880527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,3584,0.03752800159984165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,3072,0.0358551111486223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,4096,0.051517334249284535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,2560,0.029192888074451025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,3072,0.03222844335767958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,3584,0.04844355583190918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,2560,0.030812445614073012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,3072,0.04575022061665853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,2048,0.024944000773959692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,2048,0.026239999466472205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,2560,0.042747557163238525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,1536,0.01977599991692437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,1024,0.02012800011369917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,1536,0.027256000373098586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,2048,0.03883822096718682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,1024,0.013962666193644205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,1536,0.03436977664629618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,768,0.03654488921165466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,768,0.011198221809334226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,1024,0.03118577930662367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,768,0.018617777360810172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,512,0.008432000047630733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,512,0.016208888755904306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,256,0.006970666348934174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,512,0.028404444456100464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,65536,0.4121111234029134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,256,0.013212444053755866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,128,0.0033653332955307434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,32,0.003072888900836309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,32,0.01311822235584259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,256,0.023708444502618577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,128,0.013001778059535556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,16384,64,0.004053333153327306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,16384,128,0.02307111190425025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,16384,64,0.013615111509958902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,16384,0.13757867283291286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,65536,0.29650311999850804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,16384,0.08107999960581462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,65536,0.5349973572625054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,12288,0.10591733455657959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,12288,0.06515733400980632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,10240,0.0860826638009813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,10240,0.06340266598595513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,8192,0.06948000192642212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,8192,0.05371822251213921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,7168,0.06113688813315498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,12288,0.09093422359890407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,7168,0.04408177733421326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,10240,0.078274667263031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,6144,0.05341688791910807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,8192,0.06696266598171659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,6144,0.040235555834240384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,5120,0.04484000139766269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,5120,0.036497778362698026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,7168,0.06512622038523357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,4096,0.03651466634538438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,6144,0.05715733104281955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,4096,0.031826668315463595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,3584,0.03233066532346938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,3584,0.03001333276430766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,5120,0.05222400029500326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,3072,0.02822222312291463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,4096,0.04671466681692335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,3072,0.027687110834651526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,3584,0.04417066772778829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,2560,0.0240666667620341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,2560,0.025543111893865798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,3072,0.04188977678616842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,2048,0.022312889496485393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,16384,0.19391289022233751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,2048,0.023010666171709698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,2560,0.03912355502446493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,1536,0.015505777464972602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,1536,0.020967110991477966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,2048,0.034656001461876765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,1024,0.029888000753190782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,1024,0.01958400011062622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,1536,0.0321688883834415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,65536,0.34911200735304093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,1024,0.018997333116001554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,768,0.009529777699046666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,768,0.017840888765123155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,512,0.007368889119890001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,512,0.019871110717455547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,768,0.029451555675930444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,256,0.0041128889554076726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,512,0.027466666367318895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,256,0.013065777719020844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,64,0.013576000101036496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,128,0.0033164444482988785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,32,0.012791111237472959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,256,0.023584889041052923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,12288,128,0.012758221891191272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,64,0.003077333379122946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,12288,128,0.027109333210521277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,12288,32,0.003093333293994268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,16384,0.11325244108835857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,65536,0.2343520058525933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,16384,0.07106577687793307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,65536,0.4522017902798123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,12288,0.08693066570493911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,10240,0.04997244477272034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,12288,0.05749155415429009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,10240,0.07217866844601102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,8192,0.060119999779595264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,16384,0.09130933549669053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,7168,0.051700443029403687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,12288,0.0768702228864034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,8192,0.04288533329963684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,8192,0.05914044380187988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,7168,0.05507022142410278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,6144,0.03658488723966811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,7168,0.045062220758861966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,10240,0.06686666938993666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,6144,0.04533777634302775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,5120,0.03803911142879062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,5120,0.03315644462903341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,6144,0.051326221889919706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,4096,0.03140888942612542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,4096,0.031805333164003156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,5120,0.047445333666271634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,3584,0.027482666903071936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,4096,0.04390577806366814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,3584,0.027871999475691054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,3072,0.02459022237194909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,3072,0.0258586671617296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,3584,0.04248355494605171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,2560,0.020584889584117465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,65536,0.30484533309936523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,1536,0.013381332986884646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,3072,0.03878577881389194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,2560,0.02400622268517812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,1536,0.02127733329931895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,2048,0.01686311099264357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,2560,0.03705866800414191
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,2048,0.021881777379247878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,2048,0.033293333318498396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,1024,0.009848000274764167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,1536,0.03125244379043579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,1024,0.01819644371668498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,768,0.008251555263996124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,1024,0.029436445898479883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,768,0.017253332667880587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,512,0.00645955569214291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,768,0.028365333875020344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,512,0.014089778065681458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,256,0.0034471110751231513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,512,0.026891556051042344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,64,0.0029164445069101122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,256,0.01292711082432005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,128,0.0030951110853089225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,256,0.023499555057949487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,128,0.012675555215941535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,10240,32,0.0029324444217814338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,64,0.013355555633703867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,10240,128,0.022925333844290838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,10240,32,0.012647999657524956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,65536,0.18921421633826363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,65536,0.3547804355621338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,16384,0.08809066481060451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,12288,0.048383110099368624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,16384,0.05840622054206001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,12288,0.0673360029856364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,10240,0.05688266621695625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,10240,0.04636089007059733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,8192,0.04595466785960727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,8192,0.036941332949532404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,16384,0.07661600245369805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,12288,0.07083200083838569
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,7168,0.033987555238935686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,10240,0.05961510870191786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,7168,0.04365155431959364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,6144,0.03610400027698941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,5120,0.030320889419979517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,8192,0.052998221582836576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,6144,0.03128622306717767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,7168,0.053322666221194796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,4096,0.025116443634033203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,5120,0.028371555937661067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,6144,0.046536889341142446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,4096,0.024847111768192712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,5120,0.04355377621120877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,3584,0.02215733296341366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,3072,0.02161955502298143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,3584,0.02304711110062069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,4096,0.03978488842646281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,3072,0.01959111127588484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,65536,0.24561688635084364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,3584,0.037738667594061956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,2560,0.01676266723208957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,3072,0.03524622321128845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,2560,0.019911999503771465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,2048,0.03285155693689982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,2048,0.013845333622561561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,2560,0.03394755721092224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,2048,0.019295111298561096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,1536,0.011176000038782755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,1536,0.017147555947303772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,768,0.013145777914259167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,1024,0.008407110969225565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,1536,0.029920889271630183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,1024,0.01553066737122006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,768,0.007135111424658034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,1024,0.028710222906536523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,256,0.011504000259770287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,512,0.005689777847793367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,768,0.027179555760489568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,128,0.011360000405046674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,256,0.023250665929582384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,512,0.012206222448084088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,64,0.012103111379676394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,256,0.003444444388151169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,512,0.024154666397306655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,128,0.00301155551440186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,64,0.0028151110228565005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,8192,128,0.022795556320084467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,8192,32,0.0028640000770489373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,8192,32,0.011379555695586734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,65536,0.17248000038994682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,65536,0.30603734652201336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,16384,0.08105155494478014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,16384,0.05377599928114149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,12288,0.0649520026312934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,12288,0.0443520016140408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,10240,0.05361510978804695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,16384,0.07438844442367554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,10240,0.03926488757133484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,8192,0.04520533482233683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,10240,0.057349330849117704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,12288,0.0637635588645935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,8192,0.03453777896033393
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,8192,0.054809778928756714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,7168,0.03940355446603563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,5120,0.029973334736294214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,7168,0.03170488940344916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,5120,0.026704889204767015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,6144,0.034861332840389676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,6144,0.030905776553683813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,7168,0.04850844542185465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,6144,0.04492355717553032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,4096,0.022509333160188463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,5120,0.042394667863845825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,65536,0.2262373235490587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,4096,0.02288977801799774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,3584,0.020184889435768127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,4096,0.03854488995340135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,2560,0.015325332681337992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,3584,0.02165422174665663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,3072,0.017863111363516915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,3584,0.03647733396954007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,3072,0.02239199976126353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,3072,0.03475911087459988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,2560,0.019167999426523846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,2048,0.012878222597969903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,1536,0.01658933361371358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,2560,0.03327644533581204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,2048,0.017641777793566387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,1024,0.01538133290078905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,1536,0.01051111105415556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,2048,0.031186666753556993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,1024,0.007868444754017724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,1536,0.029476443926493328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,512,0.012147555748621622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,768,0.006647999915811751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,1024,0.027994665834638808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,768,0.012540444731712341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,512,0.0052471111218134565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,768,0.027043556173642475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,128,0.01128000020980835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,256,0.0032551110618644287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,256,0.011514667007658215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,32,0.0028275555620590844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,512,0.023707555400000677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,128,0.0029928889125585556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,256,0.023028444912698533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,7168,64,0.0028159999185138275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,65536,0.15762133068508571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,7168,128,0.022631110416518316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,64,0.01202933324707879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,7168,32,0.011317333413494958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,65536,0.26595910390218097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,16384,0.07199022505018446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,16384,0.050148444043265454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,10240,0.045533332559797496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,12288,0.05626844697528415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,12288,0.04139111108250088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,12288,0.06028266747792562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,16384,0.0718453327814738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,10240,0.03664533297220866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,8192,0.03774311145146688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,8192,0.0324853327539232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,7168,0.03315466642379761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,10240,0.054939554797278516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,7168,0.03022577696376377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,8192,0.04907466636763679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,6144,0.03346133232116699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,6144,0.0278479986720615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,7168,0.048235555489857994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,5120,0.024705777565638225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,65536,0.20067021581861708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,6144,0.04420977830886841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,5120,0.025175111161337957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,3584,0.022101332743962605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,4096,0.020743111769358318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,5120,0.04071377714474996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,4096,0.0218240006102456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,3584,0.018631110588709515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,4096,0.03733333283000522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,3072,0.01633777717749278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,3072,0.019488000207477145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,3584,0.035713778601752386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,2048,0.017224889662530687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,2560,0.014169777433077494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,3072,0.03417244553565979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,1536,0.016286222471131217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,2560,0.018391110830836825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,2048,0.01201600001917945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,2560,0.033095998896492854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,1536,0.009650666680600908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,1024,0.028090665737787884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,2048,0.03068888849682278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,1024,0.007276444799370236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,1024,0.013975110318925647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,1536,0.029225776592890423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,768,0.006240889016124938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,768,0.012479110724396177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,512,0.004804444395833545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,768,0.025511110822359722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,512,0.01201600001917945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,256,0.0032115555885765287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,512,0.023807111713621352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,256,0.011506666739781698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,128,0.003006222140457895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,256,0.023136888941129048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,128,0.011321778098742167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,64,0.0027546667390399506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,6144,128,0.023012444376945496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,64,0.011976000335481433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,6144,32,0.0028151110228565005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,16384,0.05793688694636027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,6144,32,0.011238222320874533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,65536,0.22559110323588052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,65536,0.13339377774132624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,16384,0.04452444447411431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,12288,0.04455822375085619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,12288,0.037132445308897234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,10240,0.037477334340413414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,10240,0.03365600109100342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,16384,0.06637599733140734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,12288,0.06011466847525703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,8192,0.030794666873084173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,8192,0.031201776531007554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,65536,0.17861421902974448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,10240,0.054604444238874644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,7168,0.027296889159414504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,8192,0.04888800117704603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,7168,0.027839110957251653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,6144,0.024110222856203716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,6144,0.042704890171686806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,6144,0.02516977820131514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,7168,0.04634666774008009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,5120,0.020620443754725985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,5120,0.022855111294322546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,4096,0.01715555621518029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,5120,0.039192001024881996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,3584,0.03475111060672336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,4096,0.02073777715365092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,4096,0.036377777655919395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,3584,0.01533422205183241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,3584,0.019778667224778067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,2048,0.010130667024188572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,3072,0.013585777746306526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,3072,0.018571555614471436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,2560,0.011752888560295105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,3072,0.03356444504525926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,2560,0.01788444485929277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,2560,0.032320890161726214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,2048,0.016567111015319824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,1024,0.012707555459605323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,1536,0.008721777962313758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,2048,0.03029688861634996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,1024,0.02591200007332696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,768,0.01239999963177575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,1536,0.015346666177113851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,1024,0.006707555717892117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,1536,0.028971556160185073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,768,0.005668444352017508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,512,0.003730666720204883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,768,0.025785777303907607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,512,0.011911110745535957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,512,0.02402488887310028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,256,0.003105777833196852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,256,0.011546666423479715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,128,0.0029057777590221832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,256,0.023123555713229712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,128,0.0112631113992797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,64,0.0026826666047175727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,5120,128,0.022664000590642292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,64,0.01202311118443807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,5120,32,0.0026977778308921387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,5120,32,0.011153777440388998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,65536,0.10920355717341106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,65536,0.17747288280063203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,16384,0.04846310946676466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,16384,0.08156444629033406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,16384,0.03908533189031813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,12288,0.0384844442208608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,12288,0.03305600086847941
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,10240,0.03275022241804335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,10240,0.02954044606950548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,7168,0.024499555428822834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,12288,0.05579288800557455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,8192,0.02718044486310747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,8192,0.026239999466472205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,65536,0.17050133811102974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,6144,0.022640888889630634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,5120,0.019336000084877014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,10240,0.06874222225613065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,8192,0.04575555523236593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,7168,0.024296889702479046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,6144,0.02204533252451155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,7168,0.04487199915779961
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,6144,0.04058488872316148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,5120,0.020789333515697055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,4096,0.016327111257447135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,5120,0.037598222494125366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,4096,0.01943288909064399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,3584,0.012797333300113678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,4096,0.03537599907981025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,3584,0.018519111805491976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,3072,0.011527111132939657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,3584,0.03408888975779215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,3072,0.017872888180944655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,2560,0.01012711144155926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,3072,0.03276266654332479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,2560,0.016872889465755887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,2048,0.00865955568022198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,1536,0.028075555960337322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,2560,0.03195733494228787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,2048,0.01591822173860338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,1536,0.0074177781740824384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,2048,0.029473778274324205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,1536,0.01348888874053955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,1024,0.005967999911970562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,1024,0.012484444512261285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,768,0.004443555656406614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,1024,0.02532622218132019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,768,0.012119111087587146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,512,0.0034151110384199354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,768,0.025117332736651104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,512,0.011886222494973077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,256,0.0030906666070222855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,512,0.02362488872475094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,256,0.011306666665607028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,128,0.0028204443968004654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,32,0.0026728889594475427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,256,0.023235556152131822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,128,0.011182222101423474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,4096,64,0.0026746667507621977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,65536,0.09709244304233128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,4096,128,0.022760000493791368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,64,0.011877333124478659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,4096,32,0.0111191107167138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,65536,0.1567724413341946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,12288,0.030847999784681533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,16384,0.04387199878692627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,16384,0.036074668169021606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,12288,0.03401333424780104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,16384,0.060806221432156034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,10240,0.029498666524887085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,10240,0.05067200130886502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,10240,0.027834667099846735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,8192,0.024559999505678814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,12288,0.0547617773214976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,8192,0.024808888634045918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,7168,0.022304889228608873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,8192,0.04569244384765625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,7168,0.02288533416059282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,6144,0.039600001441107854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,6144,0.019871999820073444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,65536,0.16962756050957573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,7168,0.042882667647467725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,6144,0.021800888909233943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,5120,0.017457778255144756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,4096,0.03581600056754218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,3584,0.01828444500764211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,5120,0.020233778489960563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,4096,0.012772444221708508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,5120,0.037984887758890785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,4096,0.018784888916545443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,3584,0.011744889120260874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,3072,0.010581333604123859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,3072,0.017495110630989075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,3584,0.03357155455483331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,2560,0.009320889082219865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,3072,0.032638221979141235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,2560,0.016340444485346477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,2560,0.03144977821244134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,2048,0.008160889148712158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,2048,0.015584889385435315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,1024,0.01238666640387641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,1536,0.00674933319290479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,2048,0.029583109749688044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,1536,0.013057777451144325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,1024,0.005584888988071018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,1536,0.027219555444187585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,768,0.003987555702527364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,256,0.0030737777964936364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,1024,0.025156444973415796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,512,0.023432888918452795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,256,0.011418666276666852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,768,0.012059555285506778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,512,0.003428444473279847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,768,0.02419022222359975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,512,0.011880888707107969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,128,0.002805333377586471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,128,0.011180444724029966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,256,0.022987556126382615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,64,0.0026782221264309357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3584,128,0.022805333137512207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,64,0.011830222275522021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3584,32,0.0026417778184016547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3584,32,0.011057777537239922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,65536,0.13637510935465494
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,65536,0.08458933565351699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,16384,0.03889777925279405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,16384,0.032594667540656194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,12288,0.030543999539481267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,12288,0.028134223487642076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,16384,0.05903999672995674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,10240,0.02659911082850562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,12288,0.05478488736682468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,65536,0.14897421995798746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,10240,0.025429333249727886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,8192,0.022121777137120564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,10240,0.05021333363321093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,8192,0.02273511058754391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,7168,0.020075556304719713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,8192,0.04409511221779717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,7168,0.021351110604074266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,6144,0.017794667018784415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,7168,0.04191733400026957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,6144,0.02013066742155287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,6144,0.03895111216439141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,5120,0.015616888801256815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,5120,0.01901155544651879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,4096,0.014121777481502958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,4096,0.018124444617165458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,5120,0.037046220567491316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,3584,0.0129120002190272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,3584,0.03356444504525926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,4096,0.03461600012249417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,3584,0.017251556118329365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,3072,0.032412442896101214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,3072,0.011792889071835412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,2560,0.03136355678240458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,3072,0.016708445217874315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,2560,0.008763555851247575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,2560,0.01587022178702884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,2048,0.00759733302725686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,2048,0.014257777068350049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,1536,0.006435555716355641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,1536,0.013125333521101209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,2048,0.029157333903842505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,1536,0.028021332290437486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,1024,0.005593777944644292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,1024,0.012404444317022959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,768,0.004031110968854693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,1024,0.0248355550898446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,768,0.011980444192886353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,512,0.003510222252872255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,768,0.02446222139729394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,512,0.011847111086050669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,256,0.003061333257291052
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,512,0.02360444433159298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,256,0.011340444286664328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,128,0.0027706666539112725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,256,0.02279111080699497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,128,0.011200000014570026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,64,0.0026595555245876312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,3072,128,0.022646221849653456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,64,0.011879111329714457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,3072,32,0.002691555561290847
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,3072,32,0.011047110789351992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,65536,0.11580888430277507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,65536,0.07439911365509033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,16384,0.03281955586539374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,16384,0.02948444419436985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,12288,0.026320889592170715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,16384,0.05853244331147936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,12288,0.02567822237809499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,10240,0.022830223043759663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,65536,0.13605688677893743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,8192,0.043700443373786084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,12288,0.053004443645477295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,10240,0.023017777336968318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,8192,0.01925511161486308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,10240,0.04853777752982246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,8192,0.02106133268939124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,7168,0.017487110363112558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,6144,0.03864177730348375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,7168,0.02003466586271922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,6144,0.01565511193540361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,7168,0.04163555635346307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,6144,0.019121777680185106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,5120,0.013872000078360239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,5120,0.01815111107296414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,4096,0.011968000067604912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,5120,0.037477334340413414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,4096,0.017303999927308824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,3584,0.011310222248236338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,4096,0.034161776304244995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,3584,0.01682222220632765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,3072,0.010552888943089379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,3584,0.03351022137535943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,3072,0.016415999995337594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,2560,0.00812888890504837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,3072,0.03240799903869629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,2560,0.015231110983424716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,2048,0.007027555671003129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,2560,0.03128088845147027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,2048,0.0136808885468377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,1536,0.006292444550328785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,2048,0.026927999324268762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,1536,0.012823111481136747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,1024,0.004288889053795072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,1536,0.025616000096003216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,1024,0.0123831108212471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,768,0.0036915557252036203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,1024,0.024862221545643274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,768,0.012144889268610211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,512,0.003399111123548614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,768,0.024288000331984624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,512,0.011847111086050669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,256,0.0030417777597904205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,512,0.023312888211674158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,256,0.011249778171380361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,128,0.0028080000645584534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,256,0.02289155622323354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,128,0.011223110887739392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,64,0.00263288886182838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2560,128,0.022659555077552795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,64,0.01180355581972334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2560,32,0.0026293332791990707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2560,32,0.010964444114102257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,65536,0.09162133269839817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,65536,0.06708977619806926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,16384,0.027069333526823256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,16384,0.027426666683620874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,12288,0.02188266648186578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,65536,0.14351377222273085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,16384,0.05721422036488851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,8192,0.015950222810109455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,12288,0.023710222707854375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,10240,0.019127110640207928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,12288,0.051029334465662636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,10240,0.021665778425004747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,7168,0.01900533338387807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,10240,0.0465866658422682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,8192,0.019855999284320407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,6144,0.01835555500454373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,7168,0.014581332604090372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,6144,0.03828622235192193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,8192,0.04282133446799385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,6144,0.013173333472675748
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,7168,0.04145688811937968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,5120,0.012150222228633033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,5120,0.017464000317785475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,4096,0.01051644484202067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,4096,0.01646844380431705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,5120,0.035968889792760216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,3584,0.010096888575288985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,4096,0.034308443466822304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,3584,0.015895111693276297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,3072,0.009394666386975182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,3584,0.03276622295379639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,3072,0.015505777464972602
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,2560,0.007266666326257918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,3072,0.03191466795073615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,2560,0.013576889203654395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,2048,0.006665777828958299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,2560,0.03085244364208645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,2048,0.013169777890046438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,1536,0.005604444278611078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,2048,0.026111111044883728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,1536,0.012651555240154266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,1024,0.003947555604908201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,1536,0.025628444221284654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,1024,0.012299555871221753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,768,0.003653333418899112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,1024,0.024870221813519795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,768,0.01219377749496036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,512,0.0032622222271230486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,768,0.024409777588314478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,512,0.011772444678677453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,128,0.01111288865407308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,256,0.0029768887907266617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,512,0.0233324451578988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,256,0.011254222028785281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,128,0.0027715555495686005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,32,0.010909332997269101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,65536,0.07122488816579182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,256,0.02312711046801673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,64,0.002611555573013094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,2048,64,0.011623111036088733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,2048,32,0.002616000051299731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,2048,128,0.022660444180170696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,65536,0.0616497794787089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,16384,0.022104889154434204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,16384,0.025340444511837427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,12288,0.01793244481086731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,16384,0.05400088760587904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,12288,0.022188444932301838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,65536,0.13797244760725233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,10240,0.015846222639083862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,12288,0.04888355400827196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,10240,0.0207306659883923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,8192,0.013438222308953604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,10240,0.04561866654290093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,8192,0.01920977731545766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,6144,0.017684444785118103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,7168,0.0125946667459276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,5120,0.01054044481780794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,8192,0.04211911227968004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,7168,0.0184542222155465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,6144,0.011584888729784222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,7168,0.04033333394262526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,5120,0.016805332567956712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,6144,0.0373964442147149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,4096,0.009545777406957414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,5120,0.03621244430541992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,4096,0.01592088904645708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,3072,0.013831111292044321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,3584,0.009082666701740688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,4096,0.03328533305062188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,3584,0.01440000037352244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,2048,0.006231110956933763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,3072,0.008461332983440822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,3584,0.03142933381928338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,2048,0.01313244468635983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,2560,0.007072889142566257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,1536,0.004986666556861666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,1536,0.012697777814335294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,2560,0.013373333546850415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,3072,0.02942844563060337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,2560,0.028042667441897925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,2048,0.02580533259444767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,1024,0.003923555629120933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,1536,0.025440888272391424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,1024,0.012329777909649743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,768,0.003640888879696528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,1024,0.02456088860829671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,768,0.011990221838156382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,512,0.003269333392381668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,768,0.02384088933467865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,128,0.0026897777699761917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,512,0.011750222080283694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,256,0.002971555623743269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,256,0.011310222248236338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,512,0.023457777169015672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,256,0.022871999277008906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,128,0.011133333047231039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,64,0.002598222138153182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1536,128,0.02234577801492479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,65536,0.05134310987260607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,64,0.011710222396585675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1536,32,0.0026142222599850762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1536,32,0.010913777682516309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,65536,0.048278222481409706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,12288,0.020601777566803824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,16384,0.016583111551072862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,16384,0.022895110978020564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,12288,0.013540444274743399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,16384,0.04829422301716275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,65536,0.1160480048921373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,10240,0.012445333103338877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,12288,0.04613066713015238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,10240,0.019355555375417072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,8192,0.01109777804878023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,10240,0.04424355427424113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,8192,0.018146667215559218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,7168,0.010277333358923594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,8192,0.040064887868033514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,7168,0.01661688917213016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,5120,0.015441776977645027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,6144,0.009045333498054082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,7168,0.03831733266512553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,6144,0.01588888963063558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,5120,0.008813333180215623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,6144,0.03636533353063796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,4096,0.008056888977686564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,5120,0.03413866625891791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,4096,0.014280888769361707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,4096,0.029114666912290785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,2560,0.006412444429265128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,3584,0.008360889222886827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,3584,0.013926222920417786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,3584,0.029562665356530085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,3072,0.007186666958861881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,3072,0.013722666435771518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,3072,0.028262221150928076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,2560,0.013487111363146039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,2048,0.00574222207069397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,2560,0.027387556102540758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,2048,0.01293155550956726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,1536,0.00461422238085005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,768,0.003564444267087512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,2048,0.025125333004527624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,1536,0.012658666405412885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,1024,0.003956444561481476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,1536,0.0247697780529658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,1024,0.012336889074908363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,1024,0.02409244411521488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,768,0.0119982221060329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,768,0.023619555764728125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,256,0.02272533377011617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,512,0.003264888914095031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,512,0.011776000261306763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,256,0.002946666752298673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,512,0.023039110832744177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,256,0.011423110961914062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,128,0.0027262222849660446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,128,0.01109955542617374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,1024,128,0.022516444325447083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,64,0.0025519999778932994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,64,0.011719110939237805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,1024,32,0.0026026666164398193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,1024,32,0.010835555692513784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,65536,0.03831733266512553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,65536,0.04740266667471992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,16384,0.01386755539311303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,65536,0.10677422417534722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,16384,0.02163822286658817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,16384,0.04485244552294413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,12288,0.012133333418104382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,12288,0.019734222027990553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,10240,0.01093066649304496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,12288,0.04326577650176155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,10240,0.018744889232847426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,8192,0.009993777506881291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,10240,0.04163555635346307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,8192,0.017722666263580322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,8192,0.037646220790015325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,7168,0.009085333181752099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,7168,0.015972443752818637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,6144,0.008440888590282863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,7168,0.036421332094404436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,6144,0.01519466605451372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,5120,0.008190222084522247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,6144,0.03157688842879401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,5120,0.01532799998919169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,4096,0.007054222126801808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,5120,0.03051555487844679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,4096,0.014241778188281588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,3584,0.0074106670088238185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,4096,0.02854222059249878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,3584,0.013923555612564087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,3072,0.006599111275540457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,2560,0.013234666652149625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,3584,0.027985778119828966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,3072,0.013783111340469785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,2560,0.006072000082996156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,1536,0.0046302220887608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,3072,0.026994667119450037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,2048,0.025202666719754536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,2048,0.005246222019195557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,1024,0.003929777691761653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,2560,0.026616000466876562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,2048,0.012839111189047495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,1536,0.012523555921183692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,1536,0.024530667397711012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,1024,0.012231999801264869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,768,0.003630222131808599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,768,0.011950222154458364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,1024,0.023955555425749883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,512,0.0032515554792351196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,768,0.023561777340041265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,512,0.011709333293967776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,128,0.01112711098459032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,256,0.0029644444584846497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,512,0.023218666513760883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,256,0.011367999845080905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,128,0.0027191111197074256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,32,0.010896888871987661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,256,0.02273066673013899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,64,0.0025884444928831523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,768,128,0.022580444812774658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,768,64,0.011711111499203576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,65536,0.09306222200393677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,768,32,0.0026106666773557663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,65536,0.027950220637851294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,65536,0.046012444628609546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,16384,0.011094222466150919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,12288,0.040397332774268255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,16384,0.021527111530303955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,12288,0.009397333694828881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,16384,0.043062223328484424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,10240,0.03679822219742669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,12288,0.018792889184421964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,8192,0.0336444444126553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,10240,0.009099555512269339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,10240,0.01758399936887953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,8192,0.008427555362383524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,8192,0.0163137780295478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,7168,0.007726222276687622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,7168,0.015784000356992085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,6144,0.007430222299363878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,4096,0.006367110957702001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,5120,0.029327111111746892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,7168,0.03219644559754266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,6144,0.015124445160230001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,5120,0.007052444749408298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,6144,0.03033511175049676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,5120,0.014641778336630927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,4096,0.014182221558358936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,4096,0.027912888262006972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,3584,0.0069751110341813825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,3584,0.014021333720948962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,3072,0.006545777950021956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,3584,0.027448889282014634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,3072,0.01352266636159685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,3072,0.026520000563727483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,2560,0.005939555664857228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,2560,0.013395555317401886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,2560,0.026164443956481084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,2048,0.0052791109515561005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,2048,0.012963555753231049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,1536,0.004567999806669024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,2048,0.025224889318148296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,1536,0.01275111072593265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,1024,0.003930666794379552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,768,0.023333332604832117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,1024,0.012243555651770698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,1536,0.024353777368863423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,768,0.003564444267087512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,1024,0.023826667004161414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,768,0.011929777761300405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,512,0.0032791110376516977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,512,0.011706666813956367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,256,0.0029013332807355454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,128,0.011167111496130625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,64,0.002565333412753211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,512,0.022898667388492163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,256,0.011182222101423474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,256,0.022664889693260193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,128,0.0027253333893087176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,512,32,0.00258933338854048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,512,128,0.0224906669722663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,64,0.011753777662913004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,512,32,0.01093155559566286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,65536,0.017464889420403373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,65536,0.044440888696246676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,16384,0.008239111138714684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,65536,0.09133422374725342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,16384,0.02001066671477424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,12288,0.03761511047681173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,12288,0.007429333196745978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,16384,0.04163644380039639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,12288,0.017830222845077515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,10240,0.007134222322040134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,10240,0.01699733402993944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,8192,0.006656888872385025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,8192,0.016100444727473788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,10240,0.03494044476085239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,7168,0.0064106666379504735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,8192,0.03302933441268073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,7168,0.015637333194414776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,6144,0.006239111224810283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,7168,0.03166044420666165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,6144,0.015023999743991427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,5120,0.006603555546866522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,6144,0.030319998661677044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,5120,0.014503111441930136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,5120,0.02922044528855218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,4096,0.00628266649113761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,4096,0.014167111780908374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,3072,0.01366666704416275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,3584,0.006708444406588872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,4096,0.02776622275511424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,3584,0.013941332697868347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,3072,0.0063031112982167145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,3584,0.027218666341569688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,2560,0.005896889087226655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,3072,0.026668444275856018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,2560,0.013295110729005603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,2048,0.0052435555391841466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,2048,0.012960889273219638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,2560,0.02601066728432973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,1024,0.012123555772834353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,1536,0.004629333400064045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,2048,0.024803555674023096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,1536,0.012552888856993781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,1024,0.003922666526503033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,1536,0.02423200011253357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,768,0.0035662220584021676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,768,0.011781333221329583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,1024,0.02346844474474589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,512,0.0032302221904198327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,128,0.002736888825893402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,512,0.011731555892361535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,768,0.02330133318901062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,256,0.0028737777223189673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,512,0.022908444205919903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,256,0.01124622258875105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,256,0.022462222311231825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,32,0.010774222513039907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,65536,0.012276444170210095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,128,0.011119999819331698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,64,0.002551111082235972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,256,128,0.02241777711444431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,256,64,0.01165244479974111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,12288,0.006256888724035687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,256,32,0.0025920000755124623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,65536,0.04382933179537455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,16384,0.006704000135262807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,65536,0.09133955505159165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,16384,0.019496889577971566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,16384,0.039847112364239164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,12288,0.01774666706720988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,10240,0.0064399999876817065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,12288,0.03740888833999634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,10240,0.016738666428460013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,8192,0.00620266670982043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,10240,0.03511733478969998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,8192,0.01588622232278188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,7168,0.006217777729034424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,8192,0.033096889654795326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,7168,0.015455111861228943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,6144,0.005966222120655908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,7168,0.031586666901906334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,6144,0.01495733360449473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,5120,0.006311111152172089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,6144,0.030243555704752605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,5120,0.01440622243616316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,5120,0.02936089038848877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,4096,0.005929777605666055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,4096,0.013927110367351107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,3584,0.006396444307433234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,4096,0.027639110883076985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,3584,0.013515555196338229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,2560,0.013006221916940479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,3072,0.006128888991143968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,3584,0.027087110612127516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,3072,0.013304000099500021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,2560,0.005623999983072281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,3072,0.026429333620601233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,2048,0.005022222383154763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,2560,0.025973333252800837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,2048,0.012803555362754397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,1536,0.004422222160630756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,2048,0.02474222249454922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,1024,0.023629332582155865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,1536,0.012324444121784635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,1024,0.0037671112351947357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,1536,0.024159111910396155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,1024,0.011885333392355176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,768,0.0034382223255104492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,768,0.01182044463025199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,512,0.022709333234363135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,768,0.023287110858493384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,512,0.003149333306484752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,512,0.011589333415031433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,256,0.0028533333291610083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,256,0.011186666786670685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,256,0.022628444764349196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,128,0.0026844443960322272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,128,0.011060444845093621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,8,128,128,0.022474666436513264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,64,0.0025475554996066624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,64,0.011651555697123209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,128,32,0.002543111021320025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,128,32,0.010836444795131683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,65536,0.009278222090668148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,65536,0.043234666188557945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,16384,0.006081777728266186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,16384,0.01928622192806668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,12288,0.006024888820118374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,12288,0.017412443955739338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,10240,0.006255110932721033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,10240,0.016383111476898193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,8192,0.006128888991143968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,8192,0.015736889508035447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,7168,0.005960888746711943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,7168,0.01516266663869222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,6144,0.005896889087226655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,3584,0.006350222147173352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,6144,0.014582221706708273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,5120,0.00629066675901413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,5120,0.014075555735164218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,2560,0.00553688903649648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,4096,0.0058622223635514574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,4096,0.013723555538389417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,3584,0.013422222601042854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,3072,0.006028444402747684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,1536,0.01221333361334271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,3072,0.013044444223244986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,768,0.0034142221427626083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,2560,0.012800888882742988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,2048,0.004910222358173794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,2048,0.012536889149083031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,1536,0.00434933313065105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,1024,0.0037671112351947357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,1024,0.011963555382357704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,512,0.0031395554542541504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,768,0.011570666399266986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,512,0.011539555258221097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,256,0.002789333255754577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,256,0.011205332974592844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,128,0.0026319999661710528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,65536,0.008054222497675154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,128,0.011142222417725457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,64,0.0024862223201327855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,64,0.01166044423977534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,64,32,0.0024968888610601425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,64,32,0.01075111081202825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,65536,0.042174223396513194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,16384,0.006105777704053455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,16384,0.019310222731696237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,12288,0.006148444281684027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,12288,0.017445334129863314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,10240,0.006220444622966979
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,10240,0.016568000117937725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,8192,0.00609422226746877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,8192,0.015896888242827523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,5120,0.01424088908566369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,7168,0.005991999887757831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,7168,0.015199111567603217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,6144,0.005932444499598608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,3584,0.013404444687896304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,3072,0.005984888722499211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,6144,0.014908444550302295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,5120,0.006265777680608962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,4096,0.005848000033034219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,3584,0.006269333263238271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,4096,0.013761777844693927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,2560,0.005523555394675996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,3072,0.013095999757448832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,2560,0.01276622215906779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,2048,0.004920888692140579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,2048,0.012548444171746572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,1536,0.004362666772471534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,1536,0.012294222083356647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,1024,0.0037128888070583344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,512,0.011559999651379056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,1024,0.011984888878133563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,768,0.003395555540919304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,768,0.011730666789743634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,512,0.003160888950030009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,256,0.00282222218811512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,128,0.002630222174856398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,256,0.011214222345087262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,32,0.010767999622556897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,128,0.011115555961926779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,64,0.0024577778660588795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,8,32,32,0.002528000002106031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,8,32,64,0.01163644426398807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,16384,0.3519875473446316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,12288,0.2692568831973606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,16384,0.7132435374789767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,12288,0.5409324434068468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,10240,0.2266764375898573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,10240,0.44977331161499023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,8192,0.1840604411231147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,8192,0.364299562242296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,7168,0.3141591019100613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,7168,0.1894622246424357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,6144,0.3211502234141032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,6144,0.1647404432296753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,5120,0.1218328873316447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,5120,0.22416088316175672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,4096,0.18413600656721327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,4096,0.11460533406999375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,8192,0.23554844326443145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,6144,0.1842364470163981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,7168,0.21214577886793348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,10240,0.2829502158694797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,3584,0.1588133308622572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,3584,0.09069689114888509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,5120,0.15660977363586426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,12288,0.3393262227376302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,3072,0.13722666104634604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,3072,0.08036000198788114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,4096,0.12480355633629693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,2560,0.11465599801805283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,2560,0.0745591123898824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,2048,0.09128889110353257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,2048,0.05908444192674425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,1536,0.07047110795974731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,16384,0.4012248780992296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,1536,0.051903112067116626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,1024,0.04747822218471103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,3072,0.09381244579950969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,1024,0.040116445885764226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,3584,0.12009599473741318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,768,0.03741333219740126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,2560,0.09255555603239272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,2048,0.08064266708162096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,768,0.03542133172353109
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,1536,0.060941331916385226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,512,0.02594311038653056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,256,0.026916444301605225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,1024,0.050085332658555776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,512,0.031194667021433514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,768,0.0455075568623013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,256,0.014550222290886773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,512,0.040510220660103693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,128,0.008857777549160851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,128,0.02556444373395708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,256,0.03256533212131924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,64,0.006217777729034424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,65536,128,0.03006311257680257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,65536,32,0.004344888859324985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,64,0.02166755497455597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,65536,32,0.02109688851568434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,16384,0.17435288429260254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,16384,0.10039733515845405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,65536,0.35863378312852645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,12288,0.15056178304884169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,65536,0.6999315685696073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,12288,0.07964266671074761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,10240,0.10966400305430095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,10240,0.0687111086315579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,8192,0.0888062251938714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,8192,0.05771644247902764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,7168,0.0781244436899821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,7168,0.06067466735839844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,6144,0.0681555536058214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,16384,0.12972177399529353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,12288,0.1073662241299947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,10240,0.0844915575451321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,6144,0.05392711030112373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,5120,0.042168001333872475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,5120,0.05707910988065931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,8192,0.07986844248241849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,7168,0.06715110937754314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,4096,0.04640444450908237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,4096,0.03653511073854234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,6144,0.06203555398517185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,3584,0.040954666005240545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,5120,0.05642311440573799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,3584,0.034128000338872276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,4096,0.05159999926884969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,3072,0.03570844398604499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,3072,0.03168177935812209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,3584,0.04760266674889458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,2560,0.03026755650838216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,2560,0.028972443607118394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,2048,0.025978666212823655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,3072,0.04442666636572944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,2048,0.024830222129821777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,1536,0.022638221581776936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,2560,0.04172177778349983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,1536,0.01958311100800832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,2048,0.03799644443723891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,1024,0.013942221800486246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,1024,0.01977244516213735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,1536,0.03402222196261088
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,768,0.011164444188276926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,1024,0.031214220656288996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,768,0.01828444500764211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,512,0.008389333056079017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,768,0.02975733412636651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,512,0.016907556189431083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,256,0.005856888989607493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,512,0.028267555766635474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,256,0.013209777573744455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,128,0.0033697777738173804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,65536,0.40871556599934894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,256,0.02404533326625824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,32,0.012130666938092975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,128,0.013536888692114087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,64,0.0030826667530669104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,16384,128,0.023744889431529578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,16384,64,0.012977778083748288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,16384,0.136936002307468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,65536,0.2778044541676839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,16384,32,0.0030657777355776895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,12288,0.10507910781436497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,65536,0.5653235647413466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,16384,0.08020977841483222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,12288,0.06416622135374281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,10240,0.08556266625722249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,8192,0.0676177806324429
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,10240,0.05597155623965793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,8192,0.04756177796257866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,16384,0.10727644628948635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,7168,0.05961955255932278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,7168,0.04351466563012865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,10240,0.07805066638522677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,8192,0.06708888875113593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,6144,0.03973689013057285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,6144,0.05260533425543043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,12288,0.11524621645609538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,5120,0.05111555589569939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,7168,0.06362133555942111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,4096,0.03347999850908915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,5120,0.03558399942186143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,4096,0.0359208881855011
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,6144,0.05734755595525106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,3072,0.027791112661361694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,4096,0.045810666349199086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,5120,0.05311110946867201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,3584,0.03238666719860501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,3584,0.030379556947284277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,3072,0.027484445108307734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,2560,0.023749333288934495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,3584,0.04619822237226698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,2560,0.025192888246642217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,3072,0.042020443412992686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,2048,0.01993777851263682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,2048,0.036181333992216326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,2048,0.022805333137512207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,2560,0.038418667183982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,1536,0.015426667200194465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,1536,0.020582222276263766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,1024,0.011477332976129321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,1024,0.018785778019163344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,1536,0.03287377622392442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,65536,0.3570062319437663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,768,0.009517333573765224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,1024,0.029851555824279785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,768,0.017606221967273288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,512,0.007324444750944774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,768,0.028974221812354192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,256,0.02351377738846673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,512,0.015011555618709989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,256,0.004320888883537716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,512,0.02737422287464142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,64,0.012554667062229581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,256,0.012989333106411828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,128,0.0033528889632887314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,128,0.013461333182122974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,64,0.0030942221896515954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,12288,32,0.003083555648724238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,12288,128,0.023735110958417256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,12288,32,0.011756444142924415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,65536,0.2302853266398112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,12288,0.08592000272538926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,16384,0.1131333377626207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,65536,0.4416240056355794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,16384,0.07002666923734877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,12288,0.0602088901731703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,10240,0.07568266656663683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,10240,0.05554577708244324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,8192,0.0582266648610433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,16384,0.0933484435081482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,8192,0.04277333286073473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,7168,0.051479111115137734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,12288,0.07257866859436035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,7168,0.03932977716128031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,6144,0.045349333021375865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,10240,0.06537066565619574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,6144,0.03615466753641764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,8192,0.057479110029008657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,5120,0.03785422113206651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,5120,0.032828443580203585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,7168,0.061843554178873696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,4096,0.029148442877663508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,6144,0.050761779149373375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,4096,0.03059555424584283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,5120,0.04655999938646952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,3584,0.027746667464574177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,3584,0.04145688811937968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,3584,0.02741866641574436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,4096,0.04287733303176033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,3072,0.02390044430891673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,3072,0.0251519994603263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,2560,0.02033866610791948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,2560,0.023551111419995625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,65536,0.29513777626885307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,3072,0.03820800118976169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,1536,0.01978844404220581
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,2048,0.01647555496957567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,2048,0.021644444929228887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,2560,0.03512622250450982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,1536,0.013288000391589271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,2048,0.03314044409328037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,1024,0.009961778091059791
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,1024,0.01796444422668881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,1536,0.032256888018714056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,512,0.014314666390419006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,768,0.008202667037645975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,1024,0.029046220911873713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,512,0.025528000460730657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,128,0.0031146665828095544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,256,0.013188444077968597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,768,0.016984888248973422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,512,0.006523555351628198
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,768,0.028338667419221666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,32,0.0028942223224374983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,256,0.0035706667436493766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,256,0.023423110445340473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,128,0.013260444833172692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,10240,64,0.0029288888391521242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,10240,128,0.02362755603260464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,65536,0.18686755498250326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,65536,0.3538995583852132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,64,0.012488889197508493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,12288,0.06813244687186347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,10240,32,0.011735999749766456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,16384,0.08818044265111287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,16384,0.05793866846296522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,12288,0.047918223672442965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,10240,0.056314667065938316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,12288,0.06586844391292997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,10240,0.04184177849027845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,16384,0.07969155576494005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,8192,0.04592355423503452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,8192,0.04020177655749851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,10240,0.058924443191952176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,7168,0.040623999304241605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,7168,0.033815112378862165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,6144,0.03602577911482917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,8192,0.0514008899529775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,6144,0.030960000223583643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,7168,0.04879822333653768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,5120,0.03022222386466132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,6144,0.04569155640072293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,5120,0.028185778194003638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,4096,0.024619556135601465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,3584,0.02271822260485755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,4096,0.02478933334350586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,65536,0.24487998750474718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,5120,0.05233600073390537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,3584,0.022071111533376906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,3072,0.034415112601386175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,4096,0.039055999782350324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,3072,0.019490665859646268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,3584,0.03844088978237576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,3072,0.021220443977249995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,2560,0.01679022279050615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,2560,0.02088088956144121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,2048,0.013944000005722046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,2560,0.03284710976812575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,2048,0.01830044388771057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,1536,0.011180444724029966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,768,0.007144889069928064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,2048,0.03153422143724229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,1536,0.017014222012625802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,1024,0.008489777644475302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,1536,0.031249778138266668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,1024,0.01568177839120229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,1024,0.02814133299721612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,256,0.0033777778347333274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,768,0.013254221942689685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,512,0.005730666634109285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,768,0.027473777532577515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,512,0.012090666426552666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,512,0.02442933287885454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,256,0.011516444385051727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,128,0.003027555636233754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,256,0.023383999864260357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,128,0.011988444460762871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,32,0.010463111102581024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,64,0.0028693332440323303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,8192,128,0.023427555958429973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,8192,64,0.0112631113992797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,8192,32,0.0028453332682450614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,16384,0.05307999915546841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,65536,0.1721653276019626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,65536,0.3052168952094184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,16384,0.08408266968197292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,12288,0.06276710828145345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,10240,0.061736000908745654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,12288,0.04404088854789734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,10240,0.039834668238957725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,16384,0.07509689198599921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,8192,0.04364533225695292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,12288,0.0628124475479126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,8192,0.033891555335786604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,7168,0.038932445976469256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,7168,0.0314684444003635
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,10240,0.05705777804056803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,6144,0.03458044595188565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,8192,0.05920444594489204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,6144,0.028894222444958154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,7168,0.04746577805942959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,5120,0.029531555043326482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,5120,0.02644888891114129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,6144,0.0445493327246772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,4096,0.022837332553333704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,65536,0.2208942174911499
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,5120,0.04111999935574002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,4096,0.02272622287273407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,3584,0.02010933392577701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,4096,0.03765955567359924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,3584,0.021391999390390184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,3072,0.017883555756674874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,3584,0.0372613337304857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,3072,0.021534222695562575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,2560,0.015290665957662793
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,3072,0.033366223176320396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,2560,0.01904622217019399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,2048,0.012835555606418185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,2560,0.032285332679748535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,2048,0.017846221725145977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,1024,0.015120888749758402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,1536,0.010421333213647207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,2048,0.031095110707812842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,1024,0.02792177928818597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,768,0.01241422196229299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,1536,0.01648977730009291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,1024,0.007894222107198503
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,1536,0.030143999391131934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,768,0.006704888823959563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,256,0.011505777637163797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,512,0.0052871112194326185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,512,0.0120328888297081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,768,0.02646488944689433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,256,0.0032248888164758682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,512,0.023873777853118047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,128,0.003018666679660479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,256,0.023128888673252527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,128,0.011915555430783166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,64,0.0028355556229750314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,7168,128,0.023373333944214716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,7168,32,0.002991111121243901
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,16384,0.07109066512849596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,64,0.011141333315107556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,7168,32,0.010428444378905825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,65536,0.15709243880377874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,65536,0.2653680112626818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,16384,0.05000533329115974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,12288,0.05549777878655327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,12288,0.041159109936820135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,10240,0.04689155684577095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,16384,0.07074844174914889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,10240,0.04095910986264547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,8192,0.037598222494125366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,12288,0.05964889129002889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,8192,0.032101333141326904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,10240,0.05450222227308485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,7168,0.03288711110750834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,7168,0.030748443471060857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,8192,0.04792355497678121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,6144,0.029548443026012842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,6144,0.04319466816054451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,7168,0.04571733209821913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,6144,0.02738222148683336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,5120,0.024375110864639282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,5120,0.025399110383457605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,65536,0.21109155813852945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,4096,0.020589333441522386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,5120,0.039975110027525164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,4096,0.02198933396074507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,3584,0.018391110830836825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,4096,0.035827553934521146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,3584,0.020989333589871723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,3072,0.016275554895401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,3072,0.019488889310095046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,3584,0.03706044289800856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,2560,0.013993778162532382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,3072,0.03290844294759963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,2560,0.01830666595035129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,2048,0.012126222252845764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,2560,0.032271110349231295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,2048,0.01721155477894677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,1536,0.009738666315873465
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,2048,0.030879110097885132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,1536,0.0162444445821974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,1024,0.007336888876226213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,1536,0.030060443613264296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,512,0.004278222305907143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,1024,0.013977777626779346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,768,0.006218666831652324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,1024,0.02755555510520935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,768,0.012277333272827996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,512,0.02369955513212416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,768,0.025332444243960913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,512,0.011775111158688864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,256,0.003248888792263137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,256,0.011400000088744693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,128,0.002965333354141977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,256,0.023154666026433308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,32,0.010464000205198923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,128,0.011758222348160215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,64,0.003015111097031169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,6144,128,0.023524444964196947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,16384,0.057167112827301025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,6144,64,0.011166222393512726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,6144,32,0.0027786667148272195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,65536,0.22509866290622285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,12288,0.0366960002316369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,65536,0.13240088356865778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,16384,0.04393511017163595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,12288,0.04403911034266154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,16384,0.06849777698516846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,10240,0.03769777880774604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,8192,0.029525332980685767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,10240,0.032794667614830866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,12288,0.05754844347635905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,7168,0.027448000179396734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,8192,0.030421333180533514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,10240,0.05237599876191881
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,7168,0.031381332212024264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,8192,0.04722399844063652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,6144,0.024010666542583044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,6144,0.02481066683928172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,7168,0.045280890332327954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,5120,0.020432889461517334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,4096,0.020392000675201416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,6144,0.041570666763517596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,5120,0.022218666142887537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,4096,0.01699822147687276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,3072,0.013628444737858243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,5120,0.039134220944510564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,3584,0.01531022290388743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,65536,0.256569782892863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,3072,0.03244977858331468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,4096,0.0347324444188012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,3584,0.019394665956497192
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,3584,0.03603644503487481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,3072,0.01846222248342302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,2560,0.01182933317290412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,2560,0.01742755538887448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,2048,0.009933333430025313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,2560,0.03572266631656223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,2048,0.016513778103722464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,1024,0.012718222207493253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,1536,0.008456000023418004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,2048,0.030212445391549006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,1536,0.014999111493428549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,512,0.003593777616818746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,1024,0.006603555546866522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,1536,0.02935911218325297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,768,0.005373333477311664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,512,0.023873777853118047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,1024,0.028640889459186133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,768,0.01221422188811832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,768,0.02458222210407257
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,512,0.011764444410800934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,256,0.00309244439833694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,256,0.011332444846630096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,128,0.0028808888875775864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,256,0.022900443938043382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,32,0.010308444499969482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,128,0.011940444509188334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,64,0.0027066665805048416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,5120,128,0.02347911066479153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,5120,64,0.01108977778090371
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,5120,32,0.0026826666047175727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,16384,0.03851911094453599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,65536,0.17647643884023032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,65536,0.10783288876215617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,12288,0.03275200062327915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,16384,0.0484168893761105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,16384,0.06398399670918782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,12288,0.038226667377683855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,10240,0.03243466549449497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,10240,0.029366221692827012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,8192,0.02639111048645443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,12288,0.05581511391533745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,10240,0.04931911163859897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,8192,0.026957333087921143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,8192,0.04520533482233683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,7168,0.0243520008193122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,7168,0.023789332972632513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,6144,0.021781333618693884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,6144,0.022097777989175584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,7168,0.04238666759596931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,5120,0.019169777631759644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,6144,0.03956444395913018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,5120,0.020581333173645865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,4096,0.015994666351212394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,65536,0.23678223292032877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,5120,0.037992888026767306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,4096,0.018894222047593858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,3584,0.012835555606418185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,4096,0.03385688861211141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,3584,0.01815022197034624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,3072,0.012792888614866467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,3584,0.035511109564039446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,3072,0.01752000053723653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,2560,0.010145777629481422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,3072,0.03197066651450263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,2560,0.016695110334290397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,2560,0.030936890178256567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,1024,0.005913777897755305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,2048,0.00868444475862715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,2048,0.016146666473812528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,1536,0.007452444069915348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,2048,0.030244443151685927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,1536,0.013139555851618448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,512,0.00342399999499321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,1536,0.029341333442264136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,1024,0.012339555554919772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,768,0.004508444418509801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,1024,0.02460977766248915
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,768,0.012070222033394707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,768,0.023980443676312763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,512,0.011671110987663269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,256,0.00309599998096625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,512,0.023668444818920557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,256,0.011284444067213269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,128,0.0028080000645584534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,256,0.0229404436217414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,32,0.010272888673676385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,128,0.011877333124478659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,64,0.002642666714058982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,4096,128,0.023336889015303716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,4096,32,0.0026586666289303037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,4096,64,0.011018666956159802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,65536,0.15525600645277235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,65536,0.09535021914376153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,12288,0.03003644280963474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,16384,0.04336622357368469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,16384,0.035628444618648954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,12288,0.035008000002966985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,16384,0.06396266486909655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,10240,0.0291955570379893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,10240,0.027707555227809485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,12288,0.054309331708484225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,8192,0.02441155579355028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,65536,0.16834044456481934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,8192,0.024259555670950148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,10240,0.05877244472503662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,6144,0.021077333225144282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,7168,0.022064000368118286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,8192,0.0443484452035692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,7168,0.022779555784331426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,6144,0.019720888800091214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,7168,0.04166577921973334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,5120,0.017821333474583097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,6144,0.03877066572507223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,5120,0.019681778219011094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,4096,0.012890666723251343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,5120,0.03708800011210971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,3584,0.034444444709353976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,3072,0.017079111602571275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,4096,0.018942221999168396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,3584,0.011744889120260874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,2560,0.01628800067636702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,4096,0.033548444509506226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,3584,0.018053332964579265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,3072,0.010629333555698395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,2560,0.009370666411187913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,3072,0.03179555469089084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,2048,0.008160889148712158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,2560,0.03080000148879157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,2048,0.015255999233987598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,1536,0.006993778049945831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,2048,0.029693331983354356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,1536,0.013256889250543384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,1024,0.005699555493063397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,768,0.011974222130245633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,1536,0.028667555914984807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,1024,0.012447111308574677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,768,0.004060444318585926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,256,0.003104000041882197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,1024,0.02444711161984338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,512,0.00342755557762252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,512,0.01165422217713462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,768,0.023893333143658105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,512,0.023571555813153584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,64,0.0026604444202449587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,128,0.02345955537425147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,256,0.011247999966144562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,64,0.010968888799349466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,128,0.0028124445428450904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3584,256,0.02313866714636485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,128,0.011723555624485016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3584,32,0.0026817777090602448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3584,32,0.010216000179449717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,65536,0.13610400093926325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,65536,0.08299822277492948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,16384,0.038521776596705117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,16384,0.032369779215918645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,12288,0.030232889784706965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,12288,0.02831377916865879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,16384,0.062163558271196157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,10240,0.02622755534119076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,10240,0.02504088812404209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,12288,0.051960001389185585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,8192,0.021912889348136053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,8192,0.0233608881632487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,10240,0.0476746658484141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,7168,0.01982399986849891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,65536,0.15842310587565103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,8192,0.04246933261553446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,7168,0.02088711162408193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,5120,0.019124444988038804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,6144,0.01757066614098019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,7168,0.040927112102508545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,6144,0.020038222273190815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,5120,0.015512888630231222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,6144,0.038197331958346896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,4096,0.033033778270085655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,4096,0.014089778065681458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,5120,0.03646399908595615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,4096,0.017844445175594754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,3584,0.01295111080010732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,3584,0.017416000366210938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,3072,0.011785777906576792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,3584,0.03425155414475335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,3072,0.016699555847379897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,2048,0.014364444547229342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,2560,0.008551999926567078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,3072,0.031231999397277832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,1536,0.012736000120639801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,2560,0.015860444969601102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,2048,0.0075839997993575195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,2560,0.030533333619435627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,1536,0.006649777707126405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,2048,0.029296000798543293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,768,0.011967110964987012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,512,0.0034248888906505373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,1024,0.005157333281305102
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,512,0.01164533363448249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,1536,0.02795377704832289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,1024,0.012303111453851065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,768,0.0037271111375755737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,1024,0.02440266642305586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,768,0.02397777802414364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,256,0.003031111011902491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,512,0.0236151119073232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,256,0.011166222393512726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,128,0.0027955555253558685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,256,0.022919111781650122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,32,0.010279999838935005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,128,0.011821333732869891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,64,0.002630222174856398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,65536,0.07363466421763103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,3072,128,0.023239110906918842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,3072,64,0.010965333216720156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,3072,32,0.00267999991774559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,65536,0.11521066559685601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,16384,0.03266133202446832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,16384,0.02959377898110284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,12288,0.026127111580636766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,10240,0.022729777627521094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,12288,0.025368889172871906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,16384,0.06128088633219401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,10240,0.02252977755334642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,12288,0.05180444320042928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,65536,0.14628799756368002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,7168,0.019951999187469482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,8192,0.01906222270594703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,10240,0.046907554070154824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,8192,0.02092355489730835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,7168,0.017450667089886136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,8192,0.04220266805754768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,6144,0.01588888963063558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,7168,0.04189599884880913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,6144,0.019144889381196763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,5120,0.013756444056828817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,6144,0.037780443827311196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,5120,0.017840888765123155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,4096,0.012106666962305704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,5120,0.03737422161632114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,4096,0.01739377776781718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,3584,0.011287111375066968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,4096,0.03283910950024923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,3584,0.01667288939158122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,3072,0.01039644413524204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,3584,0.033639109796947904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,3072,0.01611199975013733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,2560,0.008058666355080074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,2560,0.03038311004638672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,3072,0.03137333194414774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,2560,0.014873777826627096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,2048,0.0070017774899800616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,2048,0.01366488883892695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,1536,0.006080889039569431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,2048,0.028703110085593328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,1536,0.013024888932704926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,1024,0.004607111215591431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,1536,0.025923555095990498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,1024,0.01218488895230823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,768,0.0037004442678557504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,768,0.011935999823941125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,512,0.023519999451107446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,1024,0.02444177700413598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,512,0.003519111209445529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,768,0.023736889163653057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,512,0.011591999895042844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,256,0.003032888803217146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,256,0.011236444115638733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,128,0.0028106667515304354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,256,0.022737777895397607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,128,0.011704000333944956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,64,0.002603555512097147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2560,128,0.02330844435426924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,64,0.010962666736708747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,16384,0.026798221800062392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2560,32,0.0026391111314296722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2560,32,0.010234666367371878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,65536,0.09145866500006782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,65536,0.06674666537178887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,16384,0.027492443720499676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,12288,0.02160355614291297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,12288,0.04922222097714742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,12288,0.023053333163261414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,16384,0.05986577934688992
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,10240,0.01902044481701321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,10240,0.022057778305477563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,65536,0.1564942200978597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,8192,0.015799999237060547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,10240,0.045293334457609385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,8192,0.0203457772731781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,7168,0.014537778165605334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,8192,0.041335999965667725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,7168,0.01924533314175076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,6144,0.013206221991115145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,7168,0.03989333245489333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,6144,0.018326222896575928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,5120,0.012198222180207571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,6144,0.037792887952592634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,5120,0.017501332693629794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,4096,0.010753778119881948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,5120,0.03626311156484816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,4096,0.01677688956260681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,3584,0.010113777385817634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,4096,0.03309244579739041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,3584,0.015978667471143935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,3072,0.00942755573325687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,3584,0.034175998634762235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,3072,0.014295111099878946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,2048,0.013222222526868185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,2560,0.007369777394665613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,3072,0.03096533483929104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,2560,0.013585777746306526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,2048,0.006591111007663939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,2560,0.028419554233551025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,1536,0.0053591111467944244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,2048,0.026915555198987324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,1536,0.012616888516479068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,1024,0.004238222208287981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,1536,0.026269333230124578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,1024,0.012159110771285163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,768,0.0036115555299652945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,1024,0.02428177826934391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,768,0.01202222208182017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,512,0.0032915555768542816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,768,0.023797333240509033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,256,0.022861333356963262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,512,0.011517333487669626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,256,0.002951111023624738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,512,0.02346666653951009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,256,0.01125600023402108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,64,0.010899555351999072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,128,0.0027386666172080566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,128,0.01166488892502255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,64,0.002611555573013094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,2048,128,0.023443554838498432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,2048,32,0.0025920000755124623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,2048,32,0.010246222217877706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,65536,0.07058222426308526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,65536,0.060749332110087075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,16384,0.02197511163022783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,16384,0.0259244441986084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,12288,0.01736177835199568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,12288,0.04825333423084683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,10240,0.020888888173633154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,16384,0.0568808913230896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,12288,0.02222844461599986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,65536,0.14500178231133357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,10240,0.015743111570676167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,8192,0.013390222357379066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,10240,0.044459554884168834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,8192,0.019500444332758587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,7168,0.012607110871209038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,8192,0.04029511080847846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,7168,0.018748443987634447
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,6144,0.011596444580290051
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,7168,0.03919822308752272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,6144,0.017818666166729398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,5120,0.010512000156773461
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,6144,0.03693422343995836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,5120,0.01682222220632765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,4096,0.01016088906261656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,5120,0.03555644551912943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,4096,0.015026667051845126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,3072,0.013824000126785703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,3584,0.008953777452309927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,4096,0.031662222411897444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,3584,0.014051554931534661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,3072,0.008560000194443597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,3584,0.032749334971110024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,2560,0.007088888850477006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,3072,0.028246220615175035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,2560,0.013352000051074557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,1536,0.012655110822783576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,2048,0.006170666466156642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,2560,0.02753422160943349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,2048,0.012913777596420713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,768,0.0035982223020659555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,2048,0.026384888423813715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,1536,0.004951111144489712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,1024,0.003929777691761653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,1536,0.025581333372328017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,1024,0.01220444424284829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,1024,0.024146666129430134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,768,0.011711999773979187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,512,0.0032835555159383347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,768,0.02381422287887997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,512,0.011595555477672152
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,256,0.0029582221888833572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,512,0.023315555519527856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,256,0.011241777903503843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,128,0.0027182222240500977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,256,0.022821333673265245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,32,0.010144000252087912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,128,0.01168444421556261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1536,128,0.02313066687848833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,64,0.0026062221990691293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1536,64,0.01090311093462838
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1536,32,0.002587555597225825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,65536,0.047967112726635404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,65536,0.04975110954708523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,16384,0.016048888365427654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,16384,0.022477333744366963
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,12288,0.01346933344999949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,16384,0.05228444602754381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,8192,0.011101333631409539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,65536,0.12116177876790364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,12288,0.020032889313167997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,7168,0.01018311083316803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,10240,0.012239110966523489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,7168,0.017439110411538016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,12288,0.04666399955749512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,10240,0.0191928893327713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,6144,0.01650399963061015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,10240,0.04284711016549004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,8192,0.017997332745128207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,8192,0.03863111138343811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,6144,0.009220444493823582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,7168,0.03842488924662272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,5120,0.008774222599135505
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,6144,0.03548622131347656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,5120,0.015146666102939181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,4096,0.008303999900817871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,5120,0.034466667307747736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,4096,0.0140666663646698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,3584,0.008088889221350351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,2560,0.00654666663871871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,4096,0.028460443019866943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,3584,0.013674666484196981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,3072,0.007220444579919179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,3584,0.029935999049080744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,3072,0.013542222479979197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,3072,0.027608889672491286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,1536,0.004615111069546806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,2560,0.013279111021094851
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,2560,0.026425777210129633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,2048,0.005486222190989389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,2048,0.012867555850081973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,2048,0.02546399997340308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,1536,0.01258399999803967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,1024,0.003948444293604957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,1536,0.025295999315049913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,1024,0.01218933363755544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,768,0.003536000019974179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,1024,0.02383822202682495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,768,0.01178311142656538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,512,0.003302222324742211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,256,0.011145778000354767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,768,0.023499555057949487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,512,0.011575111084514193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,256,0.0029475556479560006
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,512,0.022973333795865376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,128,0.002744888886809349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,32,0.002921777880854077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,256,0.022611555125978258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,128,0.01165600038237042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,1024,64,0.002921777880854077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,1024,128,0.023178666830062866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,64,0.01092977739042706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,1024,32,0.0101742222905159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,65536,0.037604444556766085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,65536,0.04569777846336365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,16384,0.013801777528391944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,16384,0.021653332644038733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,65536,0.11116088761223687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,12288,0.011819555527634092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,16384,0.04893689023123848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,12288,0.019732443822754752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,10240,0.010821333361996545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,12288,0.04355555441644457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,10240,0.01854222185081906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,8192,0.009755555954244401
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,10240,0.04056444433000352
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,8192,0.017321777012613084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,7168,0.008921778036488427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,8192,0.03695822093221877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,6144,0.0313297775056627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,7168,0.015628443823920358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,6144,0.008384889198674096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,4096,0.014030221435758801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,7168,0.03467822074890137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,6144,0.015036443869272867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,5120,0.00814577771557702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,5120,0.014815110299322339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,5120,0.03112799922625224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,4096,0.007409777906205919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,3584,0.0070977773931291364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,4096,0.02788177794880337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,3584,0.013782222237851886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,3584,0.02892711096339756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,3072,0.006506666541099548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,2048,0.00527555536892679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,3072,0.013651555611027611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,3072,0.02631644407908122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,2560,0.006215999937719769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,2560,0.013167999684810638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,2560,0.025674665967623394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,2048,0.012799999780125089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,1536,0.004602666530344221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,2048,0.025311999850802954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,1536,0.01258044441541036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,1024,0.003924444317817688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,1536,0.02479466630352868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,1024,0.012080888781282636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,768,0.0035911111368073356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,1024,0.023710222707854375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,768,0.011799111134476133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,256,0.011160000330872007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,768,0.02331466641690996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,128,0.0028906667398081887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,512,0.003301333429084884
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,512,0.011513777905040316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,256,0.0032053333189752367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,512,0.02290222214327918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,256,0.022646221849653456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,128,0.011676444775528379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,64,0.002733333243264092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,768,128,0.023072888453801472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,64,0.010866666833559671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,768,32,0.002720000015364753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,16384,0.021319111188252766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,768,32,0.010100444157918295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,65536,0.02758666707409753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,65536,0.04350489046838549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,10240,0.00906044410334693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,16384,0.01107022249036365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,65536,0.10034755865732829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,8192,0.008489777644475302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,12288,0.009844444692134857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,16384,0.043340444564819336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,12288,0.01811022228664822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,12288,0.04007644454638163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,10240,0.016912000046836007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,10240,0.03577333357599046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,8192,0.015839111473825242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,7168,0.007910221815109253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,8192,0.03208444515864054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,7168,0.015390222271283468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,6144,0.006984000404675801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,7168,0.031045334206687078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,6144,0.014903111590279473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,5120,0.00739733295308219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,6144,0.02975999977853563
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,5120,0.014499555031458536
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,4096,0.006181333214044571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,5120,0.028917332490285236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,4096,0.013956444131003486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,3584,0.006867555694447623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,4096,0.026722666290071275
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,3584,0.013781333135233985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,3072,0.006294222341643439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,3584,0.027640889088312786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,3072,0.013450666434235044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,2560,0.006054222169849608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,3072,0.02555111050605774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,2560,0.013144888811641268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,2048,0.005342222336265776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,2560,0.025054223007626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,2048,0.01293777757220798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,1536,0.004722666823201709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,768,0.003669333126809862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,2048,0.024692444337738886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,1536,0.012615111139085559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,1024,0.004003555410438114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,1536,0.02435644467671712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,512,0.011680000358157687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,1024,0.012085333466529846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,1024,0.023408000667889912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,768,0.011818666425016193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,512,0.0033395555284288195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,256,0.022595556245909795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,768,0.02316355539692773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,256,0.003024000053604444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,512,0.022841778066423204
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,256,0.011222222612963783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,128,0.0027644443843099806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,128,0.01163999984661738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,64,0.0026657777941889236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,512,128,0.023026666707462732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,64,0.010879110958841113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,512,32,0.0026355555488003623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,16384,0.02021422154373593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,512,32,0.010211555494202508
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,65536,0.016865778300497267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,65536,0.04311644368701511
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,16384,0.008389333056079017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,65536,0.09044977691438462
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,12288,0.007246221933099959
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,16384,0.04068444503678216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,12288,0.017679111825095285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,12288,0.036121779017978244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,10240,0.007273777491516537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,10240,0.01686311099264357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,8192,0.0063973334100511335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,10240,0.03410222133000692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,8192,0.01593155496650272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,7168,0.006490666833188798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,5120,0.006552000012662675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,8192,0.03162755568822225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,7168,0.01540177729394701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,6144,0.006281777802440856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,7168,0.030432889858881634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,6144,0.01480888823668162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,6144,0.02918400035964118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,5120,0.01424088908566369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,3584,0.013776000175211163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,4096,0.006167999986145232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,5120,0.028194665908813477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,3584,0.026104888982243005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,4096,0.013968888256284924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,3584,0.006591111007663939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,4096,0.026524444421132404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,3072,0.006240889016124938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,3072,0.01346933344999949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,2560,0.005934222290913264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,3072,0.025555556019147236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,2560,0.013069333301650153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,2048,0.00527555536892679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,2560,0.02492977844344245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,2048,0.012800888882742988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,1536,0.004596444467703502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,768,0.003567999849716822
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,2048,0.024498666326204937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,1536,0.012512889173295764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,1024,0.0038977778620190094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,1536,0.02409066590997908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,1024,0.012183999849690331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,1024,0.023224888576401606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,768,0.011771555576059552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,512,0.0032142222755485107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,768,0.02304711110062069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,512,0.011592888997660743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,256,0.0029279999434947968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,512,0.022912888063324824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,256,0.011208888557222156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,128,0.023123555713229712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,128,0.0026968889352348116
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,256,256,0.022451554735501606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,128,0.01162755572133594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,64,0.002579555536309878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,64,0.010760000182522668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,256,32,0.002579555536309878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,256,32,0.010061333576838175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,65536,0.011552888486120435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,65536,0.04272088739607069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,16384,0.006424888968467712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,65536,0.09014044205347697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,16384,0.019310222731696237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,12288,0.006404444575309753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,10240,0.0166640000210868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,16384,0.040397332774268255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,8192,0.00620266670982043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,12288,0.017493334081437852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,10240,0.0063884444534778595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,12288,0.03577422102292379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,8192,0.01578044394652049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,10240,0.033886220720079206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,7168,0.006149333384301927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,8192,0.03143999973932902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,7168,0.015283554792404175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,6144,0.005984000033802456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,7168,0.030185778935750324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,6144,0.014648889501889547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,5120,0.006298666612969504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,6144,0.029110223054885864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,5120,0.014179555906189812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,4096,0.005904888941182031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,4096,0.013883555928866068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,5120,0.02851288848453098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,3072,0.013083555632167392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,3584,0.006355555521117316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,4096,0.02660088903374142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,2560,0.0127688886390792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,3584,0.013452444639470843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,3072,0.006130666782458623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,3584,0.026135111848513286
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,2560,0.005591111050711737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,3072,0.025381333298153345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,2048,0.005014222115278244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,2560,0.02493511140346527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,2048,0.012501333322789935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,1536,0.004411555412742826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,1536,0.01221422188811832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,2048,0.024271999796231587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,1024,0.003726222034957674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,1536,0.02401422295305464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,1024,0.011868444581826528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,768,0.00351555562681622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,1024,0.02313422163327535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,768,0.011750222080283694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,512,0.0032675556010670135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,768,0.023035556077957153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,512,0.011363555159833698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,256,0.0029057777590221832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,512,0.022836445106400385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,256,0.011149333582984077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,128,0.002752888947725296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,256,0.022376888328128394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,32,0.010141332944234213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,128,0.011608888705571493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,64,0.00253866674999396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,4,128,128,0.023136888941129048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,128,64,0.010764444039927589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,128,32,0.002573333266708586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,65536,0.009631111390060848
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,16384,0.005960000058015187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,10240,0.016186666157510545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,65536,0.041999111572901406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,8192,0.015415110521846347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,16384,0.019073777728610568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,12288,0.006017777654859755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,12288,0.017151110702090792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,10240,0.006216889040337668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,8192,0.006191111273235745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,7168,0.005967111223273807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,7168,0.015003555350833468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,6144,0.005850666513045629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,6144,0.014367111855083041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,5120,0.006278222219811545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,5120,0.013717333475748697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,4096,0.005877333382765452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,4096,0.01364177796575758
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,2048,0.004927111168702443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,2560,0.012660444610648684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,3584,0.006311999840868845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,3584,0.013267555170589022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,3072,0.005970666805903117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,3072,0.013003555436929068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,2560,0.005547555370463266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,2048,0.012526222401195101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,1536,0.004361777669853634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,1024,0.003689777933888965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,1536,0.0120328888297081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,768,0.0034053333931499054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,1024,0.011834666960769229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,256,0.011112000379297467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,768,0.011583111352390714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,64,0.0025519999778932994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,512,0.003140444349911478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,32,0.002558222247494592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,256,0.002855111120475663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,65536,0.008224888808197444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,32,0.010079111489984725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,512,0.011346666349305047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,64,128,0.0026871110830042097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,128,0.011591999895042844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,64,64,0.010841777755154504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,16384,0.006131555471155379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,65536,0.040608889526791044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,12288,0.006082666830884085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,16384,0.01922044489118788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,12288,0.01734755602147844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,10240,0.006181333214044571
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,10240,0.01646844380431705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,8192,0.006085333310895496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,8192,0.01536622146765391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,7168,0.005977777971161737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,5120,0.0140657772620519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,7168,0.015089778436554803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,6144,0.0058551111982928375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,5120,0.006268444574541516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,6144,0.014567999376191033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,4096,0.005842666659090254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,3584,0.006294222341643439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,4096,0.013541333377361298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,3584,0.013240888714790344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,3072,0.006001777946949005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,2560,0.00554666668176651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,3072,0.01297511077589459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,2560,0.012736000120639801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,1024,0.011801777614487542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,2048,0.004880889008442561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,2048,0.01240888900227017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,1536,0.004351110921965705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,512,0.011230222052998014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,1536,0.012068444656001197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,1024,0.0037288889288902283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,768,0.0034311111602518293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,512,0.003129777808984121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,768,0.011589333415031433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,64,0.01074577785200543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,256,0.0028106667515304354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,128,0.0025902222841978073
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,256,0.01109955542617374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,128,0.011620444556077322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,64,0.0025191110455327565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,4,32,32,0.002569777684079276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,4,32,32,0.010018666585286459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,16384,0.3502675427330865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,12288,0.2686328887939453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,16384,0.7107111083136665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,12288,0.5392693413628472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,10240,0.22486488024393717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,10240,0.4491644435458713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,8192,0.36274311277601456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,8192,0.21357422404819063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,7168,0.1832266648610433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,7168,0.31374933984544545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,6144,0.2704079945882161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,6144,0.14196266068352595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,5120,0.22489688131544325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,5120,0.12387643920050727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,4096,0.18352088663313124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,4096,0.09992088874181111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,8192,0.2316373454199897
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,6144,0.18097866906060112
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,7168,0.20907999409569633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,10240,0.28372621536254883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,3584,0.08977066808276707
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,5120,0.1574880017174615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,3584,0.1584977838728163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,3072,0.13683288627200657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,3072,0.0794186658329434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,4096,0.13361067242092556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,2560,0.1141128937403361
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,2560,0.07818844583299425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,2048,0.09092533588409424
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,2048,0.06498933500713773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,1536,0.07040711243947347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,1536,0.04855377806557549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,12288,0.34217511283026797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,3584,0.11811821990542942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,3072,0.0987573332256741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,16384,0.40298043357001406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,1024,0.04726577798525492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,2560,0.08558222320344712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,1024,0.051402668158213295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,1024,0.04478666517469618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,2048,0.07663910918765597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,768,0.03695999913745456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,1536,0.06288266844219632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,768,0.036342223485310875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,512,0.02552533315287696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,512,0.03191466795073615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,768,0.046310222811169095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,256,0.014327110515700446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,256,0.03499199946721395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,512,0.039741333987977766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,256,0.028049776951471966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,128,0.008811555802822113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,128,0.0247688889503479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,64,0.006275555739800136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,65536,128,0.02871733241611057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,64,0.0211191111140781
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,65536,32,0.004372444417741564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,65536,32,0.020817778176731534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,16384,0.17401956187354195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,65536,0.357697778277927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,16384,0.09963466723759969
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,12288,0.1324124468697442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,65536,0.6995902061462402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,12288,0.07916000154283312
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,10240,0.10991733604007298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,10240,0.06797155406739977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,8192,0.09002221955193414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,8192,0.05703022082646688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,7168,0.0779155559009976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,7168,0.05384711093372769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,12288,0.10499111149046157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,6144,0.0680275559425354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,6144,0.047601779301961265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,16384,0.11817777156829834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,5120,0.05670399798287285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,8192,0.07560800181494819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,10240,0.09484889109929402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,5120,0.04364533225695292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,7168,0.07406044668621488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,4096,0.05184355709287855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,4096,0.036347554789649114
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,6144,0.06750133302476671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,3584,0.04118044508828057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,5120,0.05875821908315023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,3584,0.03384711013899909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,3072,0.03574400146802267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,4096,0.0513973335425059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,3072,0.03122933374510871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,3584,0.048157334327697754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,2560,0.03011288907792833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,2560,0.028628445333904688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,3072,0.052410665485594005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,2048,0.02456355591615041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,2560,0.04313866628540886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,2048,0.025219556358125474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,1536,0.019498666127522785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,2048,0.03867822223239475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,1536,0.022289777795473736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,1024,0.013891556196742587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,1024,0.019374221563339233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,1536,0.0461404456032647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,768,0.011155555645624796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,1024,0.032587554719712995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,768,0.018763555420769584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,512,0.008363555702898238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,768,0.031786666976081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,512,0.01681600014368693
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,256,0.005692444327804778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,65536,0.41532444953918457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,512,0.03176622258292304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,128,0.0222080002228419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,256,0.0136817776494556
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,128,0.003351111171974076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,16384,256,0.025610667135980394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,128,0.012638222012254926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,64,0.0030293334275484085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,16384,32,0.0032115555885765287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,64,0.01221244451072481
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,16384,32,0.011893333660231696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,16384,0.1365742286046346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,65536,0.2744657728407118
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,16384,0.07996622059080336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,65536,0.5218755404154459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,12288,0.11237155066596137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,12288,0.06444622410668267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,10240,0.084824886586931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,10240,0.056032001972198486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,8192,0.07300177547666761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,8192,0.04737600021892124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,12288,0.09422933393054539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,7168,0.06160710917578804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,7168,0.045253333118226796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,10240,0.08302222357855903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,6144,0.05243200063705444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,8192,0.06942577494515313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,6144,0.039713776773876615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,16384,0.1745226648118761
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,5120,0.043544001049465604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,7168,0.06510755750868055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,5120,0.03592000073856778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,6144,0.058727112081315785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,4096,0.0388337771097819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,4096,0.035071111387676664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,3584,0.029313776228162978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,5120,0.052580446004867554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,3584,0.03164888752831353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,4096,0.048709332942962646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,3072,0.027806222438812256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,3072,0.02714755634466807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,3584,0.04528977804713779
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,2560,0.02366311185889774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,2560,0.024774221910370722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,2048,0.022266666094462078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,3072,0.043177776866488986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,2048,0.0199297782447603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,2560,0.041132443481021456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,65536,0.3543066713545058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,1536,0.015541333291265698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,2048,0.036095112562179565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,1536,0.020390222469965618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,1024,0.011459555890825061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,1536,0.033399111694759794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,512,0.00740444411834081
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,768,0.030656887425316706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,1024,0.021300445000330608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,768,0.00945688866906696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,1024,0.031039112144046362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,768,0.01792355544037289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,512,0.01531377765867445
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,256,0.025462221768167283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,256,0.004818666726350784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,512,0.028339554866154987
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,256,0.013812444276279874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,64,0.011867555479208628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,128,0.0033599999215867785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,128,0.012517333030700684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,64,0.0031271111220121384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,12288,128,0.022064889470736187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,12288,32,0.0031111112071408164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,12288,32,0.011918221910794577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,65536,0.22828978962368437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,16384,0.1112897793451945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,65536,0.4353031052483453
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,16384,0.06948266426722209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,12288,0.08570488956239487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,12288,0.056642664803398975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,10240,0.07214399841096666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,10240,0.04888622297181023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,8192,0.05751644240485298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,16384,0.09302666452195908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,8192,0.0429004430770874
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,12288,0.07442133294211493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,7168,0.058917330371008984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,10240,0.06611200173695882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,7168,0.039015110996034406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,8192,0.05757688813739353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,6144,0.04530222217241923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,6144,0.03580977850490146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,7168,0.054288887315326266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,5120,0.03750755720668369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,5120,0.03235377868016561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,6144,0.049792889091703624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,4096,0.030597332451078627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,4096,0.028943111499150593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,5120,0.046999110115898975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,3584,0.02713777787155575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,4096,0.0433919992711809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,3072,0.023934221929974023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,3584,0.029480000336964924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,3072,0.025148444705539282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,65536,0.3002062108781603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,3584,0.04110488957828946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,2560,0.02032444377740224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,3072,0.03875911235809326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,2560,0.022983999715911016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,2048,0.016774222254753113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,2048,0.021075555019908484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,1536,0.03229599859979417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,2560,0.03685422076119317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,1536,0.013285333083735572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,2048,0.03534400131967332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,1536,0.01959911154376136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,1024,0.009858667022652095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,1024,0.01776622235774994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,768,0.008031110796663497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,768,0.018044445249769423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,1024,0.030474666092130873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,512,0.0064426664676931165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,128,0.0031395554542541504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,768,0.030596445004145306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,512,0.013550221920013428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,64,0.0028808888875775864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,256,0.0035377778112888336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,512,0.028054220808876887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,256,0.013162666724787818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,256,0.025604445073339675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,128,0.01239999963177575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,64,0.011773332953453064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,10240,128,0.022271111607551575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,10240,32,0.002867555452717675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,10240,32,0.011551111108726926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,65536,0.18668711185455322
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,65536,0.3543537722693549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,12288,0.06734044684304132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,16384,0.08772355318069458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,16384,0.05746489100986057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,10240,0.05645333396063911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,12288,0.0513644450240665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,10240,0.041377776198916964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,16384,0.08137066496743096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,8192,0.04561511013242933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,8192,0.03599822190072801
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,12288,0.07032977872424655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,7168,0.03320711188846164
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,7168,0.04052799940109253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,10240,0.06159733401404487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,8192,0.05695822503831652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,6144,0.03558044301138984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,6144,0.030518223841985066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,7168,0.04940533306863573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,5120,0.030051555898454454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,5120,0.028786665863460962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,4096,0.02414488957987891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,6144,0.045853333340750806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,4096,0.02457511093881395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,5120,0.04338755541377597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,3584,0.02197866638501485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,4096,0.03990311092800564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,3584,0.022794667217466567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,3072,0.019294222195943195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,2560,0.019718221492237516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,3584,0.03735733363363478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,3072,0.020975111259354487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,2048,0.01830755505296919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,2560,0.016622222132152982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,3072,0.03663999835650126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,2048,0.013815110756291283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,2560,0.03371999992264642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,1536,0.011223999990357293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,1536,0.016900445024172466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,2048,0.032532443602879844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,1024,0.008423999779754216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,1536,0.031276444594065346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,1024,0.015407111909654407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,768,0.007148444652557373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,65536,0.3112604353162977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,1024,0.02935733397801717
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,768,0.014008000493049622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,512,0.005694222119119432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,768,0.029601779248979356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,512,0.011861333416567909
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,256,0.003506666670242945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,512,0.02732444471783108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,256,0.01459377838505639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,128,0.0030826667530669104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,256,0.02496799992190467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,128,0.01108088923825158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,64,0.0028231110837724474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,8192,128,0.022126222650210064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,64,0.010424888796276517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,8192,32,0.0028533333291610083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,16384,0.05284711056285434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,8192,32,0.010583110981517367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,65536,0.17085244920518664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,65536,0.30411020914713544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,16384,0.08345511224534775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,12288,0.0646044413248698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,12288,0.0437102218468984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,10240,0.05270933442645603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,10240,0.03901866740650601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,16384,0.07474311192830403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,8192,0.043846223089430064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,7168,0.038591110044055514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,8192,0.03376355436113145
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,12288,0.08632266521453857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,6144,0.03937155670589871
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,10240,0.06020266479916043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,7168,0.048264887597825795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,7168,0.031344887283113264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,8192,0.0532240006658766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,6144,0.02895555562443203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,5120,0.029355555772781372
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,4096,0.022574222750133936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,5120,0.026504889130592346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,4096,0.0221413340833452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,6144,0.05534755521350437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,5120,0.041231112347708807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,65536,0.22085511684417725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,3584,0.01982311076588101
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,3584,0.021201777789327834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,4096,0.038055999411476985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,3072,0.017701332767804463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,3072,0.021335111724005804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,3584,0.036024888356526695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,2560,0.015239111251301236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,3072,0.03548266821437412
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,2560,0.02019644445843167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,2048,0.012799999780125089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,2560,0.03327022327317132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,2048,0.01753777762254079
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,1536,0.010479999913109673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,2048,0.03231111168861389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,1536,0.017103999853134155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,1024,0.007920000288221572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,1536,0.03068888849682278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,1024,0.014635556273990206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,1024,0.029100444581773546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,768,0.006674666785531574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,768,0.013463999662134381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,512,0.0053022222386466126
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,256,0.011874666644467248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,768,0.029284444120195176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,512,0.012004444168673621
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,128,0.010969777901967367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,256,0.003330666571855545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,128,0.02194044490655263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,512,0.025031111306614343
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,128,0.003044444446762403
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,32,0.010308444499969482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,7168,256,0.02510044475396474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,64,0.0028204443968004654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,7168,64,0.01036266651418474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,7168,32,0.002801777794957161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,65536,0.264643563164605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,65536,0.15669066376156277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,16384,0.07089422146479289
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,16384,0.04966844452752007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,12288,0.05538400014241537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,12288,0.0449022220240699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,10240,0.04766222172313266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,16384,0.07252622312969632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,12288,0.06548088788986206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,10240,0.03600000010596381
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,7168,0.03274755676587423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,8192,0.038442667987611555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,8192,0.031758222315046526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,8192,0.05100355545679728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,10240,0.05865777863396538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,7168,0.0499946673711141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,7168,0.02970044480429755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,6144,0.028911110427644517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,6144,0.0273671117093828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,5120,0.026943110757403906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,5120,0.03997600078582764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,4096,0.021691555778185528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,3584,0.018911111685964797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,6144,0.04414133230845133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,65536,0.20542221599155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,5120,0.024496000674035814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,4096,0.02055911057525211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,4096,0.03636622097757127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,3584,0.02045866681469811
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,2560,0.018460444278187223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,3072,0.01666844387849172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,3584,0.03530044356981913
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,3072,0.019294222195943195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,2560,0.014088888963063559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,3072,0.03615555498335097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,2048,0.012123555772834353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,2560,0.03306400113635593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,2048,0.01723022262255351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,1536,0.009683555199040307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,2048,0.03188088867399428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,1536,0.015858666764365304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,768,0.012617777619096967
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,1536,0.030477331744299993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,1024,0.007247111035717859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,512,0.011721777419249216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,1024,0.013742222554153867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,768,0.006266666783226862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,1024,0.029079112741682265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,512,0.0041244443919923575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,768,0.027031110392676458
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,256,0.003245333416594399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,512,0.02475822303030226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,256,0.011882666912343768
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,128,0.0029848888516426086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,256,0.025126222107145522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,128,0.010999999940395355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,64,0.002777777819169892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,6144,128,0.02188355558448368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,64,0.010421333213647207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,16384,0.05790222353405423
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,16384,0.0436062216758728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,6144,32,0.0027946666296985415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,6144,32,0.010194666683673859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,12288,0.03659644391801622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,65536,0.22543289926317003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,65536,0.13109778033362493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,12288,0.04452000061670939
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,10240,0.037125332487954035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,16384,0.0694613324271308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,10240,0.032447109619776406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,12288,0.05894311269124349
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,8192,0.030573334958818223
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,7168,0.02718488872051239
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,10240,0.05631822347640991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,8192,0.02913155489497715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,7168,0.02712000078625149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,8192,0.04719110992219713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,65536,0.18327111668056914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,6144,0.024043555061022442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,6144,0.024478221933046978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,7168,0.04584533307287428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,5120,0.02047644390000237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,6144,0.043567998541726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,5120,0.022200889057583276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,4096,0.019388443893856473
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,4096,0.02034311162100898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,5120,0.040258665879567467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,3584,0.01517511076397366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,4096,0.035988443427615695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,3584,0.019389332996474374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,3072,0.013537777794731988
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,2560,0.017536888519922893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,3584,0.03472800056139628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,3072,0.01818933255142636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,2560,0.011805333197116852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,3072,0.03383200036154853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,2048,0.010066666536860997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,2560,0.032668444845411516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,2048,0.0163128889269299
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,1536,0.008401778009202745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,2048,0.031173331869973078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,1024,0.0281644430425432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,1536,0.013916444447305469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,1024,0.006592888798978593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,1536,0.030159999926884968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,1024,0.012675555215941535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,768,0.005723555468850666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,768,0.012641777594884237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,512,0.003624888757864634
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,768,0.026104888982243005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,512,0.01165244479974111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,128,0.010992888775136737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,256,0.0031537777847713898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,512,0.024895111719767254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,256,0.011883555187119378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,128,0.0029102222373088202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,256,0.02479377720091078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,64,0.002752888947725296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,5120,128,0.021874666213989258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,64,0.01035199976629681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,16384,0.04794489012824165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,5120,32,0.002733333243264092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,5120,32,0.010145777629481422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,65536,0.17625689506530762
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,65536,0.10592355330785115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,16384,0.037648889753553606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,12288,0.03807377815246582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,10240,0.029353777567545574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,12288,0.03224710954560174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,10240,0.03224888775083754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,16384,0.07515911261240642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,12288,0.05621688895755344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,8192,0.02677333354949951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,10240,0.05202844407823351
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,8192,0.02581866747803158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,6144,0.02163733376397027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,7168,0.02425333360830943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,65536,0.1830222209294637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,8192,0.04520533482233683
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,7168,0.024596444434589807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,7168,0.04249511162439982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,6144,0.021835555632909138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,5120,0.019023999571800232
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,3584,0.012776888906955719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,6144,0.043382220798068576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,5120,0.021776888105604384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,4096,0.01593955523437924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,5120,0.03892533315552606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,4096,0.019093333019150626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,4096,0.034685333569844566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,3584,0.018311111463440787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,3072,0.01149777736928728
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,2560,0.03188622328970168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,2048,0.008767111433876885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,3584,0.034134222401512995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,3072,0.017557332913080852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,2560,0.010118222071064843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,1536,0.014383110735151501
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,3072,0.033004442850748696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,2560,0.016608888904253643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,2048,0.015634665886561077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,1536,0.007416889071464539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,2048,0.03067200051413642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,1024,0.005993777679072485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,1536,0.029860443539089624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,1024,0.012107555237081317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,768,0.004418666578001446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,1024,0.02755733331044515
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,768,0.012438221938080259
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,512,0.0034604445099830627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,768,0.026400888959566753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,512,0.011508444117175208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,256,0.003089777711364958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,512,0.024861334098709956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,256,0.011899555722872416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,128,0.0028444443725877335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,256,0.02492888934082455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,128,0.010971555279360877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,64,0.0026151111556424033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,4096,128,0.021969777014520433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,64,0.010303999814722273
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,4096,32,0.0026933333526055017
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,4096,32,0.010115555591053432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,65536,0.15550933943854436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,65536,0.09416888819800483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,16384,0.043064888980653554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,16384,0.03522133496072557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,12288,0.03384888834423489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,12288,0.02997866604063246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,16384,0.06297066476609972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,10240,0.02924444609218174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,10240,0.02745955520206028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,10240,0.05185955431726244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,65536,0.17066755559709337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,12288,0.05513155460357666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,8192,0.02423288921515147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,8192,0.023806222611003455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,7168,0.04193955659866333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,7168,0.021913778450753953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,7168,0.025719111164410908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,5120,0.01739377776781718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,8192,0.04420355624622769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,6144,0.019479110836982727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,6144,0.02095199955834283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,6144,0.04250133368704054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,5120,0.019482667247454327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,4096,0.012837332983811697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,5120,0.040567109982172646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,4096,0.01861422260602315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,3584,0.03359999921586778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,3584,0.011933333343929715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,4096,0.03429866830507914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,3584,0.017678222722477384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,3072,0.010593777729405297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,3072,0.01681866745154063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,2560,0.009348444640636444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,3072,0.03484266665246751
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,2560,0.01605333387851715
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,1536,0.012785777449607849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,2048,0.008170666793982187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,2560,0.031622221072514854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,2048,0.015255111787054272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,768,0.004548444516128964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,1536,0.007172444628344641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,2048,0.03127110997835795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,1024,0.00591555568906996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,1536,0.027790221903059218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,512,0.011438222395049201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,1024,0.012148444851239523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,1024,0.025610667135980394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,768,0.012513777448071374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,768,0.025959110922283594
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,512,0.0034675554682811103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,256,0.0031146665828095544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,512,0.024526221884621516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,256,0.011734222372372946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,128,0.0028133332315418455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,256,0.024857777688238356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,128,0.010959111154079437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,64,0.0026373333401150177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3584,128,0.0217848883734809
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,64,0.010320888625250923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3584,32,0.0026648888985315957
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3584,32,0.010009778042634329
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,65536,0.1351999971601698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,65536,0.08222222328186035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,16384,0.038161777787738375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,16384,0.032046222024493747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,12288,0.029934220843844946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,12288,0.027181333965725366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,16384,0.07188977797826131
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,10240,0.025841777523358662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,10240,0.0498542222711775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,10240,0.02534488836924235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,12288,0.05410755674044291
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,8192,0.0537120004494985
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,8192,0.021923555268181696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,65536,0.15488266944885254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,8192,0.022277333670192297
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,7168,0.019670221540662978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,7168,0.020689777202076383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,6144,0.01740977830357022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,7168,0.04099999864896139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,6144,0.02256711158487532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,4096,0.017857778403494094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,5120,0.01551466683546702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,6144,0.039489779207441546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,5120,0.018771555688646104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,4096,0.013907555076811047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,5120,0.03826844361093309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,3584,0.013042666845851474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,2560,0.008652444514963362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,4096,0.03526399864090814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,2560,0.015875554747051664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,3584,0.01719555589887831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,3072,0.011552000211344825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,3584,0.0337137778600057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,3072,0.016421332955360413
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,3072,0.032650666104422674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,2048,0.0074782222509384155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,2560,0.03153333399030898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,2048,0.013769778112570444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,1536,0.006593777901596493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,2048,0.030050665140151978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,1536,0.012608889076444836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,1024,0.005256888767083486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,1536,0.028804444604449805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,1024,0.012103999654452005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,768,0.004121777911980947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,1024,0.02586222191651662
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,768,0.012295111185974546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,512,0.0034862222770849862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,768,0.025992888543340895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,512,0.01144977741771274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,256,0.003063999944263034
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,256,0.011821333732869891
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,512,0.02458488941192627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,128,0.0028080000645584534
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,256,0.024801777468787298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,128,0.010880889164076911
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,32,0.010078222387366826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,64,0.0026382222357723448
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,3072,128,0.021664000219768945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,3072,64,0.010308444499969482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,3072,32,0.002675555646419525
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,16384,0.029348442951838177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,65536,0.11499644650353326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,65536,0.07291911045710246
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,16384,0.06153688828150431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,16384,0.032270222902297974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,12288,0.02843466732237074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,8192,0.018906666172875296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,12288,0.025641777449183997
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,8192,0.020831111404630873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,10240,0.022434666752815247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,12288,0.05199644300672743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,10240,0.02269688910908169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,65536,0.14207911491394043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,10240,0.05742577711741129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,7168,0.017325333423084684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,8192,0.04576177729500664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,7168,0.01979022224744161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,6144,0.015488889482286243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,7168,0.04209333327081469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,6144,0.018753778603341844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,5120,0.01422755585776435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,6144,0.03976000017589993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,5120,0.017802667286660936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,4096,0.012181333369678922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,5120,0.03839377893341912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,4096,0.0169813334941864
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,3584,0.01126844435930252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,2560,0.008082666330867344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,4096,0.034953776333067156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,3584,0.0166311115026474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,3072,0.010459555519951714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,3584,0.03319199879964193
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,3072,0.016327111257447135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,3072,0.03244355652067397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,2560,0.014780445231331719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,1536,0.012776888906955719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,2048,0.007176888485749562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,2560,0.03127022253142463
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,2048,0.01347733371787601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,1536,0.006109333286682765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,2048,0.027902222341961328
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,1024,0.004439110971159405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,1536,0.027658666173617046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,1024,0.01218933363755544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,768,0.0037226668662495087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,1024,0.02572711143228743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,768,0.012486221889654795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,768,0.025820444027582806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,512,0.003415999934077263
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,512,0.011413333316644033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,256,0.003071111109521654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,64,0.0026168889469570587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,128,0.021778666310840186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,512,0.024417777856190998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,256,0.011671110987663269
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,128,0.0027911110470692315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,65536,0.09119733174641927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2560,256,0.02472266720400916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,128,0.01086400035354826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,64,0.010206222534179688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2560,32,0.002624888800912433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,12288,0.021563554803530376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2560,32,0.010030222435792288
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,65536,0.0663742224375407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,16384,0.026685333914226953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,16384,0.02747022277779049
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,16384,0.06589955753750272
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,12288,0.022871999277008906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,65536,0.21180444293551973
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,10240,0.01899111105336083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,12288,0.05162222186724345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,10240,0.021334222621387903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,7168,0.019276445110638935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,8192,0.015817777978049386
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,7168,0.04031733340687222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,10240,0.053307556443744235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,8192,0.020066666934225295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,7168,0.01481066644191742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,8192,0.04317155480384827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,6144,0.013209777573744455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,6144,0.018200889229774475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,6144,0.03802488909827338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,5120,0.012140444583363004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,5120,0.01718755563100179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,4096,0.010667555862002902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,5120,0.03786755601565043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,4096,0.01644622286160787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,3584,0.009939555492666032
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,4096,0.03520177801450094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,2560,0.00755466686354743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,3584,0.01587377819750044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,3072,0.009510222408506606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,3584,0.032467554012934365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,2048,0.013033777475357056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,3072,0.01520088811715444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,3072,0.03151022063361274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,2560,0.013979555832015144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,2048,0.006649777707126405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,2560,0.031112889448801678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,1536,0.005507555686765247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,2048,0.02864977717399597
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,1536,0.012525333298577202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,1024,0.004121777911980947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,1536,0.026544888814290363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,1024,0.012040889097584618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,768,0.0036524443162812125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,1024,0.025416000021828547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,768,0.012295111185974546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,512,0.0032915555768542816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,768,0.02602400051222907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,512,0.0114720000161065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,256,0.002959111084540685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,512,0.02440888848569658
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,256,0.011715555356608497
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,256,0.02474844455718994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,128,0.0027182222240500977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,128,0.010847110715177325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,64,0.0026106666773557663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,2048,128,0.021631111701329548
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,64,0.010144000252087912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,2048,32,0.0025804444319672054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,2048,32,0.010074666804737514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,65536,0.07032444741990831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,16384,0.056997332308027476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,65536,0.06072533130645752
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,16384,0.021722666091389124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,16384,0.025574222207069397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,10240,0.020562666985723708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,12288,0.018040888839297824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,65536,0.14142044385274252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,12288,0.022319111559126113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,10240,0.01587288909488254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,12288,0.048842665221956044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,8192,0.01333688861793942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,10240,0.04717688759167989
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,8192,0.019101333287027147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,7168,0.012640888492266336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,6144,0.03738311264250014
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,8192,0.04280266828007168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,7168,0.018428444862365723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,7168,0.03974489039844937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,6144,0.011481777661376528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,6144,0.01772799922360314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,5120,0.010558221903112201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,5120,0.016536000702116225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,4096,0.009334222310119206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,5120,0.03579466541608175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,4096,0.015600000818570455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,4096,0.034095999267366194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,3584,0.00907466643386417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,2560,0.0070879997478591064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,3584,0.014281777871979607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,3072,0.008628444539176093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,3584,0.03140711122088962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,3072,0.013904889424641928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,3072,0.029172443681293066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,2560,0.01331377774477005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,2048,0.006074666563007567
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,2560,0.028055111567179363
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,2048,0.012817777693271637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,1536,0.004641777939266629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,2048,0.027216888136333887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,1536,0.012508444488048553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,1024,0.003944000022278892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,1536,0.02632177703910404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,1024,0.0120319997270902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,768,0.0035964445107513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,1024,0.026361778378486633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,768,0.012298666768603854
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,512,0.0032764443506797156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,768,0.025736888249715168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,512,0.011401777466138205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,256,0.0029608888758553397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,512,0.024381332927280005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,64,0.0026364444444576898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,256,0.011624000138706632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,128,0.002733333243264092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,256,0.024640889631377325
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,128,0.010782221953074137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1536,128,0.02164799968401591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,64,0.010051555931568146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,16384,0.016660445266299777
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1536,32,0.002605333303411802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1536,32,0.009927111367384592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,65536,0.04793955551253425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,65536,0.049828443262312144
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,16384,0.02232088810867733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,65536,0.13144000371297201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,16384,0.053788443406422935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,12288,0.013590222431553734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,12288,0.019885333047972787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,10240,0.012295111185974546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,10240,0.01920622256067064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,12288,0.04615022076500786
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,8192,0.04028977619277106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,8192,0.010732444624106089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,10240,0.046831111113230385
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,8192,0.01788444485929277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,7168,0.01017066670788659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,7168,0.01741422216097514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,6144,0.00924800005224016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,7168,0.04102222124735514
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,6144,0.016678222351604037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,5120,0.008966222405433655
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,6144,0.03639288743336996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,5120,0.015017777681350708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,4096,0.008393777741326226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,5120,0.03415733244684007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,4096,0.014070222775141398
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,3584,0.00815644446346495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,4096,0.029708445072174072
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,3584,0.013946667313575745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,2560,0.013080889152155982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,3072,0.0072266666425599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,3584,0.029185778564876978
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,3072,0.013538666897349887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,2560,0.006404444575309753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,2048,0.026611554953787062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,3072,0.029345777299669053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,2048,0.00544088871942626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,2560,0.02717688845263587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,2048,0.012740444805887012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,768,0.003617777592606015
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,1536,0.004666666603750653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,1536,0.012395555774370829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,512,0.0032391111469931076
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,1024,0.003956444561481476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,1536,0.025613332788149517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,1024,0.011864888999197217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,1024,0.025136000580257837
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,768,0.012161778079138862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,128,0.0026871110830042097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,768,0.02566755480236477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,512,0.011494222614500256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,256,0.0029760001020299066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,512,0.024181332853105333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,256,0.011755555868148804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,256,0.024515555964575872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,128,0.010888000329335531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,64,0.0029102222373088202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,1024,128,0.021691555778185528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,64,0.010116444693671333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,1024,32,0.0029377777957253983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,1024,32,0.010061333576838175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,65536,0.03773422373665704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,65536,0.045165333482954234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,16384,0.014159111513031853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,65536,0.10923644569185044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,16384,0.021356445219781663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,12288,0.012063110868136087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,16384,0.05082755618625217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,12288,0.019157333506478202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,10240,0.010798222488827176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,10240,0.01828266680240631
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,12288,0.04422044422891405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,8192,0.009869332942697737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,10240,0.04231111208597819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,8192,0.01685333251953125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,7168,0.009217778013812171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,8192,0.0381022228135003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,6144,0.032808889945348106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,7168,0.01609244445959727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,6144,0.008226667013433244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,7168,0.038903110557132296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,6144,0.0147706667582194
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,5120,0.008377778033415476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,5120,0.01425244410832723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,4096,0.007592000067234039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,5120,0.029920889271630183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,4096,0.013883555928866068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,3584,0.007166222565703922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,4096,0.02852533260981242
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,3584,0.013626666532622443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,3584,0.028304888142479792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,3072,0.00665866666369968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,3072,0.01330844478474723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,2560,0.026346666945351496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,2560,0.005974222388532426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,3072,0.028370665179358587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,2560,0.01314755529165268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,2048,0.005292444593376583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,2048,0.012612444659074148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,1536,0.004584889031118817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,2048,0.025861332813898723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,1536,0.012280888855457306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,1024,0.003928889003064898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,1536,0.025197333759731714
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,1024,0.01204622205760744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,768,0.0035866668654812705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,1024,0.024783111280865137
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,768,0.012130666938092975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,512,0.0032524443748924467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,768,0.0251217782497406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,512,0.011352000137170156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,256,0.003144888828198115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,512,0.024671999944580927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,256,0.011640888949235281
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,128,0.0028924445311228433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,256,0.024344889654053584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,128,0.010875555376211802
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,64,0.0027431110954946945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,65536,0.04422310988108317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,64,0.010168000227875179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,768,32,0.0027146666414207886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,768,128,0.021624000536070928
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,768,32,0.010019555687904358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,65536,0.02718844513098399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,16384,0.010942222343550788
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,10240,0.009000889129108852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,65536,0.09794400135676067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,16384,0.021041777398851182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,12288,0.009537777966923183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,16384,0.0443102220694224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,12288,0.018364444375038147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,12288,0.04081955552101135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,10240,0.01702311138312022
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,7168,0.015248888068728976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,8192,0.00849511143234041
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,10240,0.03752800159984165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,8192,0.01570755574438307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,6144,0.030271109607484605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,7168,0.007684444387753804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,5120,0.014327110515700446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,8192,0.033096889654795326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,6144,0.007381333245171442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,7168,0.03460444344414605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,6144,0.014896889527638754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,5120,0.007111111448870764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,4096,0.0064284445510970215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,5120,0.028839111328125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,4096,0.013807999591032663
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,3584,0.006769777586062749
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,4096,0.02767466670937008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,3584,0.01366222235891554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,2560,0.013169777890046438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,3072,0.00639111093348927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,3584,0.027037334111001756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,2048,0.012748444245921241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,2560,0.02613688839806451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,3072,0.013288000391589271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,2560,0.00609422226746877
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,3072,0.027808000644048054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,2048,0.0053635554181204895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,1024,0.011752888560295105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,1536,0.004734222259786394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,1024,0.02459288967980279
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,2048,0.025586666332350835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,1536,0.012321777641773224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,1024,0.00396799999806616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,1536,0.025436444414986506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,768,0.003666666646798452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,768,0.012080888781282636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,512,0.0033644443998734155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,768,0.025133333272404138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,512,0.011316444310877057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,256,0.0030257778449190986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,512,0.02418577836619483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,256,0.011665778027640449
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,128,0.002777777819169892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,256,0.02439555525779724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,128,0.010858666565683154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,64,0.0026311110705137253
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,512,128,0.021715554926130507
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,64,0.010175111393133799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,512,32,0.002642666714058982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,512,32,0.010053333308961656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,65536,0.016749334004190233
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,65536,0.04279999931653341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,16384,0.008234666453467475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,65536,0.09130666653315227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,16384,0.020112888680564035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,12288,0.007238222493065729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,16384,0.040904889504114784
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,12288,0.017765333255132038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,10240,0.007078222102589077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,12288,0.03667555583847894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,7168,0.006303999986913469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,10240,0.016626665989557903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,8192,0.006636444479227066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,10240,0.034938666555616595
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,8192,0.015583111180199517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,8192,0.03156800071398417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,7168,0.015087111128701104
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,6144,0.006194666855865055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,7168,0.0306675566567315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,4096,0.006149333384301927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,6144,0.014859555496109856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,5120,0.006540444576077991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,6144,0.02959200077586704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,5120,0.01443377799457974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,3584,0.013647110925780402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,5120,0.028551111618677776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,3072,0.006342222293217977
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,4096,0.013759999639458127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,3584,0.006640000061856375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,4096,0.027493332823117573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,3072,0.013215111361609565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,3584,0.02696888811058468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,2560,0.005932444499598608
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,3072,0.02642044425010681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,2560,0.012881778180599213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,2048,0.005249777601824866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,2560,0.025831111603313025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,2048,0.012738666600651212
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,1024,0.011919111013412476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,1536,0.004596444467703502
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,2048,0.025415110919210646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,1536,0.012423111332787408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,1536,0.025019556283950806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,1024,0.003907555507289039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,768,0.0035822221802340615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,1024,0.024569777978791132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,768,0.01219555570019616
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,512,0.0032471112079090546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,512,0.011378666592968835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,768,0.024959110551410254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,256,0.0029253332565228143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,512,0.023736889163653057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,256,0.011688888900809817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,128,0.0027031112048361036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,256,0.024450666374630396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,128,0.010804444551467896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,64,0.002562666725781229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,256,128,0.021681777305073206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,64,0.010135110881593492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,256,32,0.002562666725781229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,256,32,0.009927111367384592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,65536,0.012091555529170565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,65536,0.04232177800602383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,16384,0.00657155571712388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,65536,0.09070044755935669
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,16384,0.019190222024917603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,12288,0.006129777679840724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,10240,0.03433510992262099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,16384,0.04036533170276218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,8192,0.015487111277050443
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,12288,0.017295110556814406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,10240,0.006268444574541516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,12288,0.03642400105794271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,10240,0.016474665866957772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,8192,0.006247111078765657
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,7168,0.006143110907740063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,8192,0.03146222233772278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,7168,0.015104000767072042
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,7168,0.03047200043996175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,6144,0.005978666659858491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,6144,0.014635556273990206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,5120,0.0063528890411059065
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,6144,0.029389331738154095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,5120,0.014254222313563028
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,5120,0.02850666642189026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,4096,0.0059075554211934405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,4096,0.013886222408877479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,3584,0.00637155564294921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,3072,0.026377777258555096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,4096,0.027356444133652583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,3584,0.013393777940008374
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,3584,0.026958222190539043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,3072,0.006056889063782162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,3072,0.01294844432009591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,2560,0.005598222215970357
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,1536,0.012038222617573209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,2560,0.012712889247470431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,1024,0.0037360000941488477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,1024,0.011662222445011139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,2560,0.025806221697065566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,2048,0.005055111315515307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,2048,0.012471111284361945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,2048,0.02544533378548092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,512,0.0032720000793536506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,1536,0.004392000122202767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,1536,0.025125333004527624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,768,0.0035502223504914176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,1024,0.02442844377623664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,768,0.011971555650234222
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,768,0.024938666158252295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,512,0.011289777855078379
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,256,0.0028880000528362063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,512,0.024119110571013555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,64,0.002549333290921317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,256,0.011613333390818702
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,128,0.002700444517864121
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,32,0.009938666390048133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,256,0.024259555670950148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,128,0.01072622256146537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,65536,0.04155644443300035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,2,128,128,0.02141422198878394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,128,32,0.0025475554996066624
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,128,64,0.010071111222108206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,65536,0.009302222066455418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,16384,0.005974222388532426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,16384,0.01885333326127794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,8192,0.015234667393896313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,12288,0.00600888869828648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,10240,0.006221333311663733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,12288,0.016902221573723685
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,10240,0.016040000650617812
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,8192,0.006163555714819167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,7168,0.005952000204059813
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,6144,0.0058453331391016645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,4096,0.01351377781894472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,3584,0.00629688882165485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,7168,0.014704888065656027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,6144,0.014321777555677625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,2560,0.005531555662552516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,5120,0.006265777680608962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,2048,0.004933333231343163
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,2560,0.012610666453838348
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,5120,0.013849777479966482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,4096,0.005882666756709416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,3584,0.013199999928474426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,3072,0.006057777752478917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,768,0.0034026667061779234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,1024,0.011602666642930774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,3072,0.012793777717484368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,2048,0.012359999948077731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,1536,0.004360888981156879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,1536,0.011944889194435544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,1024,0.0037120001183615792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,768,0.011904888682895236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,512,0.0031102223114834893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,512,0.011219555305110084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,256,0.00286311118139161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,256,0.011705777711338468
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,65536,0.0083155557513237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,128,0.0027013334135214486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,128,0.01074222226937612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,64,0.002568888788421949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,64,32,0.002568888788421949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,64,0.010034666293197209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,64,32,0.00999466660949919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,16384,0.006165333506133821
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,65536,0.040405333042144775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,16384,0.01905955539809333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,7168,0.005972444597217772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,12288,0.006082666830884085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,12288,0.017128888103697035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,10240,0.006162666612201267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,10240,0.016208888755904306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,8192,0.006231110956933763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,8192,0.015303999185562134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,7168,0.014932443698247274
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,6144,0.005848888721730974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,6144,0.014534221755133735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,5120,0.006263999889294307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,5120,0.014073777529928418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,2560,0.005507555686765247
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,4096,0.005856888989607493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,4096,0.013380444712109037
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,3584,0.00630044440428416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,3072,0.005980444451173146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,3584,0.013170666992664337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,3072,0.012883555557992725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,2048,0.004957333207130432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,2560,0.01259022206068039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,2048,0.012333333492279053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,1536,0.004320888883537716
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,1536,0.01200177768866221
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,1024,0.00370755543311437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,1024,0.011607999602953592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,768,0.0033946666452619764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,768,0.011877333124478659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,512,0.003079111170437601
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,512,0.01126399967405531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,256,0.002760888801680671
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,256,0.011680000358157687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,128,0.0026355555488003623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,128,0.010788444843557147
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,64,0.00254133323000537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,64,0.010144888526863521
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,2,32,32,0.0024871110088295406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,2,32,32,0.009988444546858469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,16384,0.34952534569634336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,16384,0.7115706867641873
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,12288,0.5409119923909506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,12288,0.26791466606987846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,10240,0.44984711541069883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,10240,0.22457332081264922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,8192,0.35950756072998047
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,8192,0.18194577429029676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,7168,0.31269245677524143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,7168,0.18866577413347033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,6144,0.14089777734544542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,6144,0.2997777726915148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,5120,0.22434311442905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,5120,0.12025600009494358
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,4096,0.21752444903055826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,4096,0.0995253324508667
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,8192,0.23531021012200248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,10240,0.2851795620388455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,12288,0.33802045716179746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,7168,0.20491200023227266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,16384,0.3977804448869493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,3584,0.15867643886142307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,6144,0.18098755677541098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,3584,0.09589866797129314
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,2560,0.11393511295318604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,5120,0.15601600541008845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,3072,0.13641866048177084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,3072,0.08359822299745347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,2560,0.06940088669459026
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,4096,0.1524240043428209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,2048,0.09099644422531128
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,2048,0.0576346649063958
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,1536,0.07005421982871161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,1536,0.048977779017554395
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,1024,0.05548266569773356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,1024,0.03857422206136916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,768,0.03678844372431437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,768,0.03629866573545668
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,512,0.02531555626127455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,512,0.03186755710177951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,3584,0.11742577287885879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,256,0.014316444595654806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,3072,0.10371644629372491
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,256,0.02715555495686001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,128,0.008761777646011775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,2560,0.09391999906963772
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,128,0.024263999528355066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,128,0.027829332484139338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,64,0.006238222122192383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,64,0.020925333102544148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,65536,32,0.00435288871328036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,65536,32,0.02086666723092397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,2048,0.07218133078681098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,1536,0.061287111706203885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,65536,0.3567502233717177
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,1024,0.06017155779732598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,65536,0.6995137532552084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,16384,0.09896444612079197
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,16384,0.17370666397942436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,768,0.04469777809249031
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,12288,0.13255643844604492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,12288,0.08116977744632296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,512,0.03862311111556159
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,16384,0.13902044296264648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,10240,0.10946755276785956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,10240,0.077437334590488
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,12288,0.1057022213935852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,8192,0.08869066503312852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,10240,0.08603377474678887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,8192,0.05698133177227444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,7168,0.07781066497166951
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,7168,0.06508621904585096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,8192,0.07897155814700656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,5120,0.05648710992601183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,7168,0.05124533176422119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,6144,0.06808444526460436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,4096,0.04609600040647718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,6144,0.046626667181650795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,6144,0.06145866711934408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,5120,0.04158044523662991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,5120,0.05438311232460869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,3584,0.03383200036154853
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,4096,0.03612622287538316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,3584,0.040728890233569674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,4096,0.048792888720830284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,3072,0.035424000687069364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,3072,0.030663109487957422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,3584,0.04738755689726936
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,2560,0.030014223522610132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,2560,0.041568888558281794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,3072,0.0470106667942471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,2560,0.028628445333904688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,2048,0.024471110767788355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,2048,0.03731733229425218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,1536,0.0349333319399092
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,65536,256,0.031621333625581526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,2048,0.02513511147763994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,1536,0.019331556227472093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,1536,0.02265600032276577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,1024,0.013854222165213691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,1024,0.019704000817404855
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,768,0.011128000087208219
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,1024,0.03107644451989068
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,768,0.018723555737071566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,512,0.008409777449236976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,768,0.03042222062746684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,512,0.017471111483044095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,512,0.028712888558705647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,256,0.005843555761708154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,256,0.012821333275900947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,256,0.023099554909600153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,128,0.0033617777129014335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,128,0.012078222301271228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,64,0.0030337776988744736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,128,0.021148444877730474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,65536,0.5218453407287598
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,65536,0.2741413381364611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,64,0.011967110964987012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,16384,32,0.00304888892504904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,16384,32,0.011874666644467248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,16384,65536,0.41544445355733234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,16384,0.12158666716681586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,10240,0.08659733004040188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,16384,0.13673333326975504
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,12288,0.0875742236773173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,8192,0.07196711169348823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,16384,0.0908542209201389
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,12288,0.10429688957002427
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,65536,0.3530604574415419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,12288,0.06453777684105767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,10240,0.05592978000640869
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,10240,0.0773457752333747
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,8192,0.04692799846331278
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,7168,0.0595724450217353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,8192,0.0678222245640225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,6144,0.06069866816202799
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,7168,0.04310755597220527
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,6144,0.05273600088225471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,7168,0.06150488720999824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,6144,0.0395342210928599
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,5120,0.04379555583000183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,5120,0.03528177738189697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,4096,0.03531288769510057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,3584,0.043655110730065234
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,3072,0.03261511193381415
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,5120,0.05207111106978523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,4096,0.03144266539149814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,3584,0.0317813323603736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,4096,0.04658311274316576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,3584,0.029519110918045044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,3072,0.026689777771631878
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,2560,0.023708444502618577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,3072,0.04169244567553202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,2560,0.024692444337738886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,2048,0.0196079992585712
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,2560,0.03862844573126899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,2048,0.021983999345037673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,2048,0.03480622172355652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,1024,0.02983733349376255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,1536,0.01661422186427646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,768,0.0181057784292433
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,1536,0.020745777421527438
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,1536,0.03307644526163737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,1024,0.011591110792424945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,1024,0.018204443984561496
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,768,0.009577777650621202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,768,0.029436445898479883
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,512,0.007453333172533248
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,512,0.015288889408111572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,512,0.027686221732033625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,256,0.004664888812435998
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,256,0.01260444439119763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,256,0.02256711158487532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,128,0.003359111025929451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,128,0.011739555332395764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,12288,128,0.02124711043304867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,64,0.0031102223114834893
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,64,0.011727111207114326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,12288,32,0.003105777833196852
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,12288,32,0.011607999602953592
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,65536,0.4436684449513753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,12288,0.05563910802205404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,65536,0.22771199544270834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,16384,0.1114026705423991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,16384,0.06908177667193942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,12288,0.08608888917499119
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,16384,0.08969422181447347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,10240,0.07150310940212674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,12288,0.0707368916935391
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,10240,0.05464977688259549
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,8192,0.057964444160461426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,10240,0.06501599815156725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,8192,0.04179733329349094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,65536,0.29621688524882
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,7168,0.05075200067626106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,8192,0.05556977457470364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,7168,0.038819554779264666
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,6144,0.04479111234347025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,7168,0.05371555685997009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,6144,0.03563200102912055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,5120,0.03757955630620321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,6144,0.049470222658581205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,5120,0.032335999939176775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,5120,0.04475111100408766
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,4096,0.030645334058337744
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,4096,0.02915999955601162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,3584,0.04020266731580099
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,4096,0.041251556740866765
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,3584,0.02714577813943227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,3584,0.02741777731312646
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,3072,0.02367377777894338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,3072,0.024861334098709956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,2560,0.02010222276051839
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,3072,0.037074665228525795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,2560,0.02328088879585266
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,2048,0.016718222035302054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,2560,0.037462221251593694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,2048,0.020872889293564692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,1536,0.0151555554734336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,2048,0.03256977929009332
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,1536,0.01979022224744161
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,1536,0.03141422073046366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,1024,0.009827555881606208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,1024,0.01755466726091173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,1024,0.029226667351192896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,768,0.008188444707128737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,768,0.017278222574128043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,768,0.029117332564459905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,512,0.0064195555945237475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,128,0.011696000066068439
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,512,0.014296889305114746
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,512,0.02677511175473531
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,256,0.003887999802827835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,256,0.012367111113336353
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,128,0.0031644445326593188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,256,0.02260444396071964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,64,0.002887111157178879
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,10240,128,0.021129777034123737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,64,0.011511111425028907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,10240,32,0.0028782222006056043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,10240,32,0.011566222541862063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,65536,0.353091557820638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,65536,0.1860897805955675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,16384,0.08757511112425063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,16384,0.05702222055859036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,12288,0.06712266471650866
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,16384,0.08076444599363539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,12288,0.04692622356944614
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,65536,0.2463982105255127
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,10240,0.05625155236985949
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,7168,0.04048266675737169
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,7168,0.0365937782658471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,12288,0.0673911107911004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,10240,0.04110399881998698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,8192,0.045555555158191256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,10240,0.05899022022883097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,8192,0.03565511107444763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,8192,0.050344000260035195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,6144,0.03566844595803155
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,7168,0.050160000721613564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,6144,0.03038311004638672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,6144,0.04449511236614651
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,5120,0.030062221818500098
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,5120,0.027540443672074214
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,4096,0.02456444501876831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,3072,0.019408888287014432
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,5120,0.04305244485537211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,4096,0.02421155571937561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,3072,0.03403910994529724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,2560,0.01664444473054674
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,4096,0.03797333439191183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,3584,0.02199733257293701
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,3584,0.022729777627521094
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,3584,0.03910222318437364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,3072,0.020854221449957952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,2560,0.020071999894248117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,2048,0.013763555222087435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,1536,0.030431999100579157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,1024,0.008372444245550368
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,2560,0.03304444419013129
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,2048,0.017975999249352347
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,2048,0.031187554200490315
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,1536,0.011178666518794166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,1536,0.017113778326246474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,1024,0.015262222952312894
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,768,0.007109333243634965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,256,0.0033946666452619764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,1024,0.028147555059856836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,768,0.013934221532609729
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,768,0.02838311261600918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,512,0.0058559998869895935
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,512,0.012347555822796293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,512,0.02474222249454922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,256,0.011251555548773872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,128,0.003068444422549672
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,256,0.02234755622016059
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,128,0.010346666806273991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,64,0.0028568889117903183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,8192,128,0.020989333589871723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,64,0.010263111028406356
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,8192,32,0.0028275555620590844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,8192,32,0.010288000106811523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,65536,0.3035591178470188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,65536,0.17039910952250162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,16384,0.07228889067967732
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,16384,0.08011733161078559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,16384,0.052228444152408175
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,12288,0.0620702240202162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,12288,0.04838488830460442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,10240,0.052412443690829806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,12288,0.0738275580936008
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,10240,0.0383857786655426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,65536,0.23391734229193795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,8192,0.04298222064971924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,7168,0.04584799872504341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,6144,0.03410666518741184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,10240,0.054413331879509814
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,5120,0.02903377678659227
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,8192,0.033321777979532875
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,7168,0.04454044500986735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,7168,0.03132088979085287
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,8192,0.04924177792337206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,6144,0.028407110108269587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,3584,0.01976444489426083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,6144,0.04317955507172478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,5120,0.025644444757037695
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,5120,0.04086222251256307
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,4096,0.022142221530278523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,4096,0.022325333621766832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,4096,0.03994755612479316
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,3584,0.021627555290857952
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,3584,0.03609866566128201
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,3072,0.017656889226701524
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,3072,0.019874667127927143
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,3072,0.03310399916436937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,2560,0.015234667393896313
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,2560,0.01917866700225406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,2048,0.012784889174832238
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,2560,0.032800889677471586
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,1024,0.014936000108718872
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,2048,0.017343110508388944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,2048,0.030805332793129817
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,1536,0.010421333213647207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,1536,0.016725333200560678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,1536,0.030016889174779255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,1024,0.007907555335097844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,768,0.006761777732107375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,1024,0.027929776244693335
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,768,0.013955555028385587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,768,0.027032888597912256
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,512,0.005404444618357553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,512,0.012270222107569376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,256,0.0033599999215867785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,512,0.024987555212444727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,64,0.01019377758105596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,256,0.011179555621412067
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,256,0.022325333621766832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,128,0.0030168888883458245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,128,0.01037599974208408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,7168,128,0.020946666598320007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,64,0.0027786667148272195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,7168,32,0.0028124445428450904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,16384,0.04943733414014181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,7168,32,0.010238221950001186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,65536,0.2638337877061632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,65536,0.15558132860395643
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,16384,0.07072533501519097
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,12288,0.05594399902555677
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,10240,0.03851200143496195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,16384,0.07469511032104492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,12288,0.04030311107635498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,65536,0.20573066340552437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,10240,0.04670133193333944
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,12288,0.061428447564442955
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,8192,0.03831733266512553
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,8192,0.03145688772201538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,10240,0.05274577935536703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,7168,0.032607999112870954
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,8192,0.04695733388264974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,7168,0.02937244375546773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,6144,0.02871555421087477
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,7168,0.04588355620702108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,6144,0.027225777506828308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,5120,0.025066667132907446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,6144,0.04204622242185804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,5120,0.024265777733590867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,3584,0.020792000823550753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,5120,0.03886399997605218
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,4096,0.020717332760492962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,4096,0.02144977781507704
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,3584,0.01867999964290195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,4096,0.03544355432192484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,3072,0.016140444411171805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,3584,0.03487022386656867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,3072,0.01942222151491377
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,2560,0.014293332894643148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,1536,0.009641778137948778
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,3072,0.03280266788270738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,2560,0.018406222263971966
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,2560,0.0322364436255561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,2048,0.012089778151777057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,2048,0.016686222619480558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,2048,0.03035466538535224
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,1536,0.016206221448050607
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,1536,0.030026667647891577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,1024,0.007293333609898885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,1024,0.012950222525331708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,768,0.006230222268237009
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,1024,0.027494221925735474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,768,0.01274311128589842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,512,0.004399999976158142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,256,0.022385777698622808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,128,0.002995555599530538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,768,0.025733333494928148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,512,0.012169777519173093
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,256,0.003294222263826264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,512,0.0248346659872267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,256,0.011100444528791638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,128,0.010336000058386061
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,6144,128,0.02108088963561588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,64,0.002769777758253945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,64,0.010131555298964182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,6144,32,0.0027262222849660446
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,6144,32,0.010201777848932479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,65536,0.22502666049533418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,65536,0.13068355454338923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,16384,0.05816000037723117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,16384,0.043326222234302096
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,12288,0.0436719987127516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,12288,0.03601244423124526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,16384,0.07005333238177829
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,10240,0.036912888288497925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,12288,0.05899644560284085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,10240,0.032391111056009926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,65536,0.1854399972491794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,10240,0.05397955576578776
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,8192,0.030065778228971694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,8192,0.028740445772806805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,8192,0.0452497767077552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,7168,0.026895110805829365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,7168,0.029120001527998183
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,6144,0.0239955551094479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,7168,0.04414577616585625
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,6144,0.02444444431198968
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,4096,0.02149244480662876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,5120,0.02030755579471588
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,6144,0.04054311248991225
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,5120,0.02218755582968394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,4096,0.01680888897842831
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,5120,0.03728266557057699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,3584,0.015135111080275642
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,4096,0.03470755616823832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,3584,0.019686222076416016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,3584,0.03414488832155863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,3072,0.013621333572599622
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,3072,0.01814399990770552
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,2560,0.01181333346499337
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,2048,0.029539555311203003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,3072,0.03242399957444932
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,2560,0.017774222625626456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,1024,0.0064497776329517365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,2560,0.03190755512979295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,2048,0.010199999643696679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,2048,0.016141333513789706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,1536,0.008416888614495596
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,1536,0.015792888071801927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,1536,0.029350221157073975
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,512,0.01218488895230823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,1024,0.012485333614879184
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,1024,0.027170666389995154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,768,0.00572533326016532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,768,0.012574222352769641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,512,0.003809777812825309
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,768,0.025088000628683303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,128,0.010221333139472539
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,256,0.0031831111345026228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,512,0.024235554867320593
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,256,0.010993777877754636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,256,0.022288888692855835
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,128,0.002941333378354708
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,64,0.0026977778308921387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,5120,128,0.02103288968404134
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,64,0.010151999692122141
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,5120,32,0.0028880000528362063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,5120,32,0.010087110930018956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,65536,0.1757457786136203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,65536,0.10522755649354722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,16384,0.048066665728886925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,16384,0.037652442852656044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,12288,0.03799110982153151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,12288,0.03182755576239692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,16384,0.06632533338334826
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,10240,0.03238755464553833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,12288,0.05780800183614095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,10240,0.02886311213175456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,65536,0.1758586698108249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,8192,0.02664622167746226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,10240,0.04823288983768887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,8192,0.02555022140343984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,7168,0.024300444457266066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,6144,0.03852622376547919
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,8192,0.04320089022318522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,7168,0.023542222049501207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,6144,0.02155911094612545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,7168,0.040448000033696495
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,6144,0.022052443689770166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,5120,0.018925334016482036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,5120,0.020269332660569083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,4096,0.015954666667514376
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,5120,0.03626666797531976
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,4096,0.019077334139082167
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,3584,0.012806221842765808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,4096,0.03406755460633172
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,3072,0.031574222776624895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,3584,0.018481777773963082
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,3072,0.011559999651379056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,3072,0.017440888616773818
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,3584,0.0332622230052948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,2560,0.010183999935785929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,2560,0.016785777277416654
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,2048,0.008791111409664154
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,2560,0.03136266602410211
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,2048,0.015408888459205627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,1536,0.007456888755162557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,1536,0.014524444937705994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,2048,0.02952622373898824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,1024,0.005964444329341252
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,1536,0.029126223590638902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,1024,0.011952888634469775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,512,0.01202311118443807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,768,0.004583111239804162
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,1024,0.024823110964563157
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,768,0.012502222425407834
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,512,0.0034702221552530923
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,768,0.025081778566042583
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,256,0.0030462222380770575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,512,0.024194666081004675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,256,0.01091288857989841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,32,0.002672000063790215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,128,0.002809777855873108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,256,0.02219555609756046
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,128,0.010229333407349056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,4096,64,0.0026515554636716843
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,4096,128,0.02090577781200409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,64,0.010031999813185798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,4096,32,0.01015022231472863
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,65536,0.15485333071814641
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,65536,0.09340444538328384
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,16384,0.04267466730541653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,16384,0.03515911102294922
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,12288,0.0337084432442983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,65536,0.1716568867365519
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,12288,0.02952622373898824
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,16384,0.06107644240061442
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,10240,0.028873778051800195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,12288,0.05622488922542996
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,10240,0.027328888575236004
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,8192,0.024100444383091394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,10240,0.04774577750100029
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,8192,0.024496888120969135
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,7168,0.02187999917401208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,8192,0.04193333453602261
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,7168,0.02221244408024682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,6144,0.019335110982259113
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,7168,0.039860443936453924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,6144,0.02093155516518487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,6144,0.03843555516666836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,5120,0.01718755563100179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,5120,0.01994311147265964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,3584,0.01804266704453362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,4096,0.012896888785892062
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,5120,0.03585066729121738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,3072,0.01685333251953125
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,4096,0.018269333574506972
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,3584,0.011698666546079846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,4096,0.03370133373472426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,3072,0.010574222438865237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,3584,0.03301689028739929
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,2560,0.009342222577995723
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,3072,0.031298667192459106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,2560,0.016533333394262526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,2048,0.00813333359029558
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,2560,0.0310808883772956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,2048,0.015107555521859063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,1536,0.006782222125265334
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,2048,0.02902666727701823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,1536,0.013566222455766467
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,1536,0.02684533264901903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,1024,0.005587555468082428
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,768,0.024846222665574815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,1024,0.01184977756606208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,1024,0.02442933287885454
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,768,0.003995555556482739
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,768,0.012465777496496836
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,256,0.022135999467637803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,512,0.0034755555291970572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,128,0.01016444464524587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,512,0.012071111136012606
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,256,0.0030702222138643265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,512,0.024342222346199885
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,256,0.010935111178292168
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,128,0.0028088889602157804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3584,128,0.02090666691462199
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,64,0.0026319999661710528
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,65536,0.08155377705891927
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,64,0.010071111222108206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3584,32,0.002671111168132888
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3584,32,0.010112889111042023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,65536,0.13476710849338108
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,16384,0.06485778093338013
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,16384,0.03807466559939914
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,65536,0.15371289518144396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,10240,0.02589599953757392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,16384,0.032010667853885226
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,12288,0.029919998513327703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,12288,0.027513777216275532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,12288,0.05275555451711019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,10240,0.02418222195572323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,8192,0.021672000487645466
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,10240,0.04596977763705783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,8192,0.022084444761276245
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,7168,0.019679110911157396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,7168,0.020764445265134174
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,8192,0.04318844609790378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,6144,0.017439110411538016
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,7168,0.039135999149746366
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,6144,0.019683554768562317
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,5120,0.015416888727082146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,6144,0.03738755649990506
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,5120,0.018975999620225694
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,4096,0.014117333624098035
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,3072,0.011758222348160215
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,5120,0.035445332527160645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,4096,0.017658667431937326
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,3584,0.012920888761679331
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,4096,0.033412443266974554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,3584,0.017660443981488545
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,3584,0.0326186650329166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,3072,0.016343999240133498
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,2560,0.008739555875460306
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,2048,0.013431111143694984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,3072,0.03119022316402859
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,2560,0.016286222471131217
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,1024,0.004924444274769889
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,2048,0.00742133292886946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,2560,0.030647112263573542
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,1536,0.006597333484225803
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,2048,0.02888444397184584
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,1536,0.013012444807423485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,1536,0.026330666409598455
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,1024,0.011977777712874942
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,768,0.004118222329351637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,1024,0.0244168887535731
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,768,0.012353777885437012
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,512,0.0034640000926123727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,768,0.02467022173934513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,512,0.01198933356338077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,256,0.003064888839920362
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,512,0.02405511173937056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,256,0.01091288857989841
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,128,0.0028231110837724474
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,256,0.022095999783939783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,128,0.010133333504199982
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,64,0.002623999905255106
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,3072,128,0.020807999703619216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,64,0.010029333333174387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,3072,32,0.0026773332307736077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,3072,32,0.010083555347389644
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,65536,0.1127857764561971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,65536,0.07212444146474202
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,16384,0.03251288996802436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,12288,0.05208177698983086
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,10240,0.022514666120211285
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,10240,0.022309333086013794
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,16384,0.02919644448492262
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,12288,0.025788444611761305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,12288,0.02514222264289856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,16384,0.062479999330308705
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,65536,0.14082488748762342
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,8192,0.018921777606010437
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,10240,0.044610665904151074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,8192,0.020503110355801053
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,8192,0.042860445049073964
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,7168,0.01720711092154185
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,7168,0.019828443725903828
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,6144,0.015477332803938123
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,4096,0.012034667034943899
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,7168,0.039530667993757464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,6144,0.018781332506073844
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,5120,0.013918222652541267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,6144,0.037661333878835045
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,5120,0.017645332548353408
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,3072,0.010286221901575724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,5120,0.035601778162850276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,4096,0.01698933376206292
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,4096,0.03335288829273648
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,3584,0.011176000038782755
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,3584,0.0170168893204795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,3584,0.031938665442996554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,3072,0.015767110718621146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,3072,0.031367109881507024
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,2560,0.00794222205877304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,2560,0.015136889285511441
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,2048,0.007012444237867991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,2560,0.030547555949952867
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,1024,0.004585777719815572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,2048,0.01312711089849472
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,2048,0.028417779339684382
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,1536,0.00618488879667388
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,1536,0.012978666358523898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,1536,0.02611555490228865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,1024,0.011959999799728394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,768,0.003755555384688907
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,1024,0.02427911096149021
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,768,0.012426666915416718
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,512,0.003412444351447953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,768,0.02446488870514764
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,512,0.011952888634469775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,256,0.003072000005178981
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,512,0.024010666542583044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,256,0.010861333045694562
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,128,0.0027822222974565295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,256,0.021898667017618816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,128,0.010182222558392419
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2560,128,0.020842666427294414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,64,0.002640888922744327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,64,0.010106667048401302
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2560,32,0.0026462222966882917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2560,32,0.010079111489984725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,65536,0.09086844656202529
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,65536,0.06622222397062513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,16384,0.026752000053723652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,16384,0.027239999837345544
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,12288,0.021534222695562575
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,16384,0.061218665705786816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,12288,0.023191110955344305
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,10240,0.018792000081804063
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,8192,0.01976977785428365
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,65536,0.15803999370998806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,12288,0.05182488759358724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,7168,0.019119110372331407
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,10240,0.021419554948806763
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,10240,0.0431182218922509
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,8192,0.015846222639083862
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,7168,0.014502222339312235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,8192,0.04292355643378364
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,6144,0.013178666432698568
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,7168,0.03796266516049703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,6144,0.0181368887424469
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,5120,0.012162666353914471
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,3584,0.010034666293197209
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,6144,0.03635999891493056
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,5120,0.017332444588343304
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,3072,0.009295110901196798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,4096,0.01074577785200543
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,4096,0.016420443852742512
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,5120,0.038303110334608294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,4096,0.032418668270111084
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,3584,0.016226665841208566
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,3584,0.030427555243174236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,3072,0.014344889256689282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,2560,0.00741244438621733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,3072,0.030773331721623737
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,2560,0.013824000126785703
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,2048,0.006654222392373615
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,2560,0.03030844529469808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,2048,0.012923555241690742
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,1536,0.005306666509972678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,2048,0.026320000489552815
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,1536,0.012860444684823355
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,1024,0.003924444317817688
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,1536,0.026167111264334783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,1024,0.011903111305501727
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,768,0.003640000190999773
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,1024,0.024086222052574158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,768,0.012348444097571902
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,512,0.003293333368168937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,768,0.024863999750879075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,512,0.011960888902346293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,128,0.010173333187898
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,256,0.002987555538614591
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,512,0.02424977719783783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,256,0.010830221904648675
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,128,0.0027093332674768236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,256,0.021996445126003687
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,64,0.002575111058023241
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,2048,128,0.020851555797788832
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,64,0.010087110930018956
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,2048,32,0.002570666579736604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,2048,32,0.009970666633711921
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,65536,0.06998488638136122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,65536,0.06082399686177572
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,16384,0.02163644466135237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,16384,0.024863999750879075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,12288,0.01737777723206414
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,10240,0.020446222689416673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,16384,0.05707022216584948
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,8192,0.013443555268976422
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,65536,0.13836710982852513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,12288,0.02197066611713833
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,10240,0.015634665886561077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,12288,0.04817510975731743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,8192,0.018957333432303537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,10240,0.04200355543030632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,7168,0.0125928885406918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,8192,0.041922665304607816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,7168,0.018345778187115986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,7168,0.037252446015675865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,6144,0.011429333024554782
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,6144,0.01773066653145684
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,5120,0.01052977806992001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,5120,0.016689777374267578
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,6144,0.03574666712019179
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,4096,0.009675555759006077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,5120,0.03623022304640876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,4096,0.0148026661740409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,3584,0.008941333327028487
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,4096,0.03183555603027344
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,3584,0.014688889185587565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,3072,0.00850044439236323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,3584,0.029947555727428857
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,2048,0.006088888893524806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,3072,0.014142221874660917
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,3072,0.029992000924216375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,2560,0.0070497774415545994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,2560,0.014408888088332282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,2560,0.027354665928416785
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,2048,0.01258933295806249
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,1536,0.004651555584536659
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,2048,0.025733333494928148
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,1536,0.0127697777416971
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,1024,0.003942222230964237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,1536,0.02588622272014618
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,1024,0.011735999749766456
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,1024,0.02405333353413476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,768,0.00360977773865064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,768,0.01220533334546619
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,768,0.02440799938307868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,512,0.0033048888047536216
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,512,0.011968000067604912
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,256,0.0029297777348094513
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,128,0.020768000019921195
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,512,0.024010666542583044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,32,0.0026177778426143858
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,256,0.010857777463065254
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1536,256,0.02196177840232849
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,128,0.002696000039577484
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,128,0.010088889135254754
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1536,64,0.002583111118939188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,64,0.009884444375832876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,16384,0.022221333450741235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1536,32,0.010046222143703038
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,65536,0.047704888714684375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,65536,0.04951644606060452
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,12288,0.04506044586499532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,16384,0.01635199950801002
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,12288,0.013431111143694984
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,10240,0.03977955712212457
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,65536,0.12052444616953532
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,16384,0.05168622069888645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,12288,0.01993955506218804
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,10240,0.012278222375445895
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,10240,0.01908177799648709
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,8192,0.010805333654085795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,8192,0.01779111060831282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,7168,0.01014755583471722
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,8192,0.03910488883654276
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,7168,0.017180444465743173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,7168,0.03706044289800856
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,6144,0.00923111124171151
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,6144,0.015662221444977652
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,5120,0.008960888617568547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,6144,0.035347554418775774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,5120,0.01443555619981554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,5120,0.03456000155872769
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,4096,0.008255110846625434
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,3072,0.013417777915795645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,4096,0.013944889108339945
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,3584,0.008041777544551426
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,4096,0.028988444142871436
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,3584,0.01406044430202908
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,3584,0.028143111202451918
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,3072,0.007223111059930589
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,3072,0.02789333462715149
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,2560,0.006635555376609166
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,2560,0.013379555609491138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,2048,0.0052977779673205475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,2560,0.026564444104830425
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,2048,0.012540444731712341
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,2048,0.02520711057715946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,1536,0.004618666652176115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,1536,0.01276177747382058
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,1024,0.003940444439649582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,768,0.02415200074513753
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,1536,0.025342222717073228
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,1024,0.011720000041855706
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,768,0.0035751110149754416
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,1024,0.023611555496851604
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,768,0.012249777714411417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,512,0.003277333246337043
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,512,0.011952888634469775
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,256,0.0029608888758553397
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,512,0.023694222172101338
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,64,0.0028986665937635633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,256,0.010840000377760993
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,128,0.0027271111806233725
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,256,0.02176888949341244
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,128,0.010168888502650792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,1024,128,0.020678222179412842
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,1024,32,0.0028942223224374983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,64,0.010049777726332346
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,1024,32,0.010044444766309526
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,65536,0.03759288787841797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,65536,0.04472177889611986
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,16384,0.014088888963063559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,65536,0.11071377992630005
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,16384,0.02129511038462321
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,12288,0.011795555551846823
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,16384,0.047685331768459745
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,12288,0.01927288870016734
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,10240,0.010936888555685679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,12288,0.04289244280921089
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,10240,0.01827822294500139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,8192,0.009508444203270806
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,10240,0.038592000802357994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,8192,0.016560888952679105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,8192,0.03789155681927999
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,7168,0.00903555585278405
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,7168,0.01531555586391025
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,6144,0.008328888979223039
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,7168,0.03501777847607931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,6144,0.016201777590645682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,5120,0.008234666453467475
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,6144,0.03303822212749057
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,5120,0.01418133411142561
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,5120,0.02980888883272807
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,4096,0.007240888973077138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,4096,0.013808888693650564
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,4096,0.027590221828884546
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,3584,0.007288888924651676
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,3584,0.01387466655837165
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,3072,0.006878222028414409
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,3584,0.027373333772023518
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,3072,0.013252444565296173
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,3072,0.02632177703910404
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,1536,0.004612444589535396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,2560,0.005934222290913264
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,1536,0.012691555751694573
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,2048,0.024655999408827886
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,2560,0.013329777452680798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,1024,0.011695110963450538
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,2048,0.0052844443255000645
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,2560,0.025872000389628943
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,2048,0.012448888685968188
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,512,0.0032951111594835916
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,1024,0.003923555629120933
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,1536,0.02479377720091078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,256,0.003157333367400699
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,256,0.010779555473062726
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,768,0.003571555432346132
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,128,0.0028737777223189673
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,1024,0.023437332775857713
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,768,0.012266666524940066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,768,0.024181332853105333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,512,0.011844444606039258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,512,0.02365511159102122
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,256,0.021864000293943617
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,128,0.010085333552625444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,64,0.002712888850106133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,768,128,0.020767110917303298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,64,0.010086221827401055
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,768,32,0.0027093332674768236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,768,32,0.010029333333174387
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,65536,0.02773155603143904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,16384,0.043064888980653554
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,65536,0.043340444564819336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,16384,0.01110666659143236
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,65536,0.09932266341315375
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,16384,0.020900444851981267
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,12288,0.009763555394278632
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,12288,0.01828622155719333
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,10240,0.009129777550697327
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,12288,0.038859556118647255
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,10240,0.016545777519543965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,8192,0.008450667063395182
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,10240,0.03471822208828396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,8192,0.01564088960488637
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,7168,0.007990222010347579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,8192,0.03277777632077535
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,7168,0.015064888530307345
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,6144,0.006776888751321369
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,7168,0.029459555943806965
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,6144,0.014698666003015308
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,5120,0.007449777589903937
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,6144,0.0285173323419359
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,5120,0.014135999812020196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,4096,0.006133333262470033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,3072,0.006490666833188798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,5120,0.029438220792346533
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,4096,0.013752889302041797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,3584,0.006844444407357111
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,4096,0.02622933354642656
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,3584,0.013975110318925647
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,2560,0.01331555595000585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,3584,0.026535999443795946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,3072,0.013261333107948303
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,2560,0.006086222413513396
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,3072,0.025440000825458105
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,2048,0.005375999957323074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,2560,0.02551733288500044
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,2048,0.012430222498046027
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,1536,0.004654222064548069
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,2048,0.024285333024130926
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,1536,0.012717333104875354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,1024,0.004027555386225383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,1536,0.024708444873491924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,1024,0.011722666521867117
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,768,0.003654222107595868
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,1024,0.023285332653257582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,768,0.01219911128282547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,512,0.0033902221669753394
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,768,0.024064888556798298
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,128,0.002777777819169892
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,512,0.011883555187119378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,256,0.0030293334275484085
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,512,0.023451555106374953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,256,0.010769777827792697
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,256,0.02169955604606205
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,128,0.010112889111042023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,64,0.0026275554878844153
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,512,128,0.020624000165197585
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,64,0.010058666268984476
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,512,32,0.0026186667382717133
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,16384,0.019888000355826486
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,12288,0.007290666302045186
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,65536,0.08986044592327541
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,512,32,0.009967111051082611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,65536,0.01720533271630605
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,65536,0.04240977764129639
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,16384,0.008207110895050896
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,8192,0.006552888701359431
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,16384,0.03807910945680406
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,12288,0.01757066614098019
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,12288,0.03403822249836392
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,10240,0.007012444237867991
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,10240,0.016781333420011733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,10240,0.03200622068511115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,8192,0.015478221906556023
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,7168,0.006268444574541516
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,8192,0.030086222622129653
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,7168,0.015102222561836243
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,5120,0.014106666048367819
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,6144,0.00620977787507905
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,7168,0.02905333373281691
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,6144,0.014676445060306124
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,3584,0.006599999964237213
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,6144,0.028038223584493
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,5120,0.006618666566080517
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,4096,0.006225777582989798
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,5120,0.027144889036814373
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,4096,0.013680000272062091
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,4096,0.026251556144820318
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,3584,0.013952889376216464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,3072,0.006240889016124938
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,3584,0.02639466689692603
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,2560,0.025430222352345783
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,3072,0.013078221844302284
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,2560,0.005953777581453323
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,3072,0.025136889682875738
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,2560,0.013223110801643796
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,2048,0.0052373334765434265
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,2048,0.012439111040698158
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,1536,0.004593777573770947
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,2048,0.02418400016095903
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,1536,0.012734221915404001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,1024,0.003940444439649582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,1536,0.024501333634058636
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,1024,0.011633777783976661
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,512,0.011892444557613797
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,768,0.003573333223660787
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,1024,0.02331111166212294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,768,0.012266666524940066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,128,0.00272088891102208
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,512,0.0032515554792351196
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,768,0.023781332704755995
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,64,0.0025920000755124623
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,256,0.0029377777957253983
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,512,0.02336888843112522
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,256,0.010752889017264048
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,256,0.021589333812395733
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,128,0.010085333552625444
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,256,128,0.02051199972629547
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,64,0.009972444838947719
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,256,32,0.002562666725781229
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,256,32,0.010018666585286459
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,65536,0.012119111087587146
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,65536,0.04186311033036974
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,16384,0.006265777680608962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,65536,0.08947022093666925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,16384,0.018927999668651156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,10240,0.016418667303191293
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,12288,0.006144000010357962
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,16384,0.037654221057891846
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,12288,0.017196445001496207
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,12288,0.03407466742727492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,7168,0.006151111175616582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,10240,0.0063688887490166565
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,10240,0.03196977906756931
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,8192,0.006270222365856171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,8192,0.015503111812803479
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,8192,0.029932445949978296
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,7168,0.014941333068741692
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,6144,0.005942222144868638
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,5120,0.014088888963063559
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,7168,0.029123554627100628
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,6144,0.014531556102964612
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,5120,0.0062871111763848206
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,6144,0.028240889310836792
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,4096,0.005920889063013925
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,5120,0.027111111415757075
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,4096,0.013738666971524557
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,3584,0.006341333190600078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,3072,0.012853333519564735
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,4096,0.026169778572188482
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,3584,0.013651555611027611
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,3072,0.006084444622198741
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,2560,0.025475554996066626
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,3584,0.02660088903374142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,2560,0.005580444302823808
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,3072,0.02515911062558492
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,2560,0.012864000267452665
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,1024,0.0037537775933742523
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,2048,0.005013333426581489
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,1024,0.011552888486120435
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,2048,0.012311110893885294
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,1536,0.004421333471934001
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,2048,0.024271999796231587
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,512,0.003253333270549774
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,1536,0.012553777959611682
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,1536,0.024409777588314478
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,768,0.0035626664757728577
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,1024,0.02312266661061181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,768,0.012088889049159156
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,768,0.02370311154259576
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,512,0.011749332977665795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,128,0.010092444717884064
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,256,0.002918222298224767
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,512,0.023408888114823237
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,256,0.010773333410422007
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,128,0.002687999978661537
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,256,0.021719111336602107
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,64,0.00254844439526399
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8_block,1,128,128,0.020581333173645865
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,64,0.00996977753109402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,128,32,0.002612444468670421
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,128,32,0.009927999642160203
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,10240,0.006232888748248418
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,65536,0.009304888546466827
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,10240,0.016007999579111736
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,65536,0.04052177733845181
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,16384,0.006043555421961679
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,16384,0.018612444400787354
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,12288,0.005998222364319696
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,12288,0.016892444756295946
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,8192,0.006117333140638139
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,8192,0.015201777219772339
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,7168,0.005990222096443176
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,7168,0.01462222304609087
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,4096,0.01330044451687071
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,6144,0.0058959999846087555
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,3072,0.005960000058015187
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,3584,0.013464888764752282
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,3072,0.012693333129088083
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,2560,0.005502222312821283
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,6144,0.014137778017255994
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,5120,0.006237333433495627
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,1536,0.004328000048796336
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,5120,0.013778666655222574
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,4096,0.005869333528810077
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,3584,0.006341333190600078
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,2560,0.012787555654843649
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,2048,0.004947555561860402
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,2048,0.012104888757069906
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,1536,0.012288000020715924
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,512,0.01181688904762268
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,1024,0.003678222083383136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,768,0.0034053333931499054
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,1024,0.011536888778209686
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,128,0.010110222631030613
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,768,0.01221333361334271
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,512,0.0031386667655573953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,256,0.002874666617976295
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,256,0.01075466639465756
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,128,0.0027031112048361036
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,16384,0.006218666831652324
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,64,0.0025600000388092464
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,64,32,0.002542222125662698
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,64,0.009922666682137383
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,64,32,0.009929777847396003
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,65536,0.008289777570300633
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,10240,0.01608800060219235
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,65536,0.04020444552103678
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,16384,0.01882577770286136
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,12288,0.006110222389300664
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,12288,0.017090666625234816
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,10240,0.006133333262470033
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,8192,0.006129777679840724
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,8192,0.015391111373901367
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,7168,0.005949333310127258
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,7168,0.014765333798196582
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,6144,0.0058711109062035876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,6144,0.014430221584108142
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,5120,0.006214222146405115
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,3584,0.013415999710559845
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,3072,0.0060159998635451
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,5120,0.013966222604115805
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,4096,0.005850666513045629
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,2048,0.004920888692140579
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,2560,0.012763555679056378
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,4096,0.013390222357379066
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,3584,0.006321777900060018
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,1536,0.012288889123333825
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,2560,0.005533333453867171
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,1024,0.011519110865063138
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,3072,0.012636444634861417
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,512,0.0031306667046414483
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,1536,0.004326222257481681
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,2048,0.012064889073371887
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,1024,0.003698666476541095
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,128,0.002628444383541743
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,768,0.0034675554682811103
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,768,0.012151111331250934
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,64,0.009884444375832876
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,256,0.0028124445428450904
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,512,0.011749332977665795
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,256,0.010797333386209277
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,64,0.0025253333151340485
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,128,0.010123555858929953
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,float16,1,32,32,0.0025457777082920074
VLLM,0.14.0,NVIDIA H100 80GB HBM3,gemm,vllm_default,fp8,1,32,32,0.009952888720565373
